Detach Ricky background CLI runs#104
Conversation
|
Caution Review failedPull request was closed or merged during review No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (2)
🚧 Files skipped from review as they are similar to previous changes (1)
📝 WalkthroughWalkthroughAdds detached background run support: parent spawns a detached CLI child with an env-based run-id handoff and exits; child detects the env, reattaches by starting the local run monitor in foreground; local monitor state initialization now persists and returns a running state. ChangesDetached Background Process Feature
Estimated code review effort🎯 4 (Complex) | ⏱️ ~45 minutes Poem
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Warning There were issues while running some tools. Please review the errors and either fix the tool's configuration or disable the tool if it's a critical failure. 🔧 ESLint
ESLint skipped: no ESLint configuration detected in root package.json. To enable, add Comment |
| if (cliHandoff && process.env[DETACHED_BACKGROUND_RUN_ID_ENV]) { | ||
| return runDetachedBackgroundChild(parsed, cliHandoff, deps); | ||
| } |
There was a problem hiding this comment.
🟡 Detached child env-var check bypasses --no-run and mode safety gates, forcing workflow execution
The detached child detection at cli-main.ts:1867 only checks cliHandoff && process.env[DETACHED_BACKGROUND_RUN_ID_ENV] without validating any parsed args. In contrast, the parent-side gate shouldLaunchDetachedBackgroundRun (cli-main.ts:706-714) carefully validates parsed.runRequested === true, parsed.background === true, and parsed.mode !== 'cloud'. The child path applies none of these checks.
When runDetachedBackgroundChild is entered, it calls startLocalRunMonitor → withSafeRunOptions (local-run-monitor.ts:47), which forces stageMode: 'run' regardless of the user's intent. Because the env var is set via spawn() and inherited by all descendant processes, any nested ricky invocation from within the executing workflow (e.g., ricky local --spec 'something' --no-run) would be intercepted: it builds a cliHandoff, sees the inherited env var, enters the child path, and has its --no-run silently overridden to execute the workflow.
Minimal fix: mirror the parent-side safety checks
Adding parsed.foreground (or at minimum not having parsed.background) to the condition would prevent accidental interception of nested invocations, since only the explicitly spawned child receives --foreground from backgroundChildArgs.
| if (cliHandoff && process.env[DETACHED_BACKGROUND_RUN_ID_ENV]) { | |
| return runDetachedBackgroundChild(parsed, cliHandoff, deps); | |
| } | |
| if (cliHandoff && parsed.foreground && process.env[DETACHED_BACKGROUND_RUN_ID_ENV]) { | |
| return runDetachedBackgroundChild(parsed, cliHandoff, deps); | |
| } |
Was this helpful? React with 👍 or 👎 to provide feedback.
There was a problem hiding this comment.
Addressed in bd3897c. The detached-child path now requires the explicit spawned-child shape (--run --foreground, non-cloud, not --no-run) instead of only checking the inherited env var, and runDetachedBackgroundChild deletes RICKY_DETACHED_BACKGROUND_RUN_ID before starting the monitor so descendant Ricky invocations do not inherit it. Added regression coverage for an inherited env var plus nested --no-run command, and asserted the env is cleared before local execution.
Ricky Eval ReviewRun: Passed: 8 | Needs human: 43 | Reviewable: 43 | Missing output: 0 | Failed: 0 | Skipped: 0 Human Review CasesThese cases passed deterministic checks and include captured Ricky output for a human verdict against their REVIEW
|
| Step | Action | Result |
|---|---|---|
| 1️⃣ | Parse arguments – Ricky reads --mode local and the --spec string. |
The CLI knows you want a local workflow generation based on the textual spec. |
| 2️⃣ | Validate the spec – The spec text is passed through the Spec‑Validator gate (deterministic, JSON‑schema‑based). | If the spec is syntactically valid, validation passes; otherwise you get a clear error (e.g., “Spec must be a JSON/YAML object with name, steps, …”). |
| 3️⃣ | Generate workflow contract – The Workflow‑Builder creates a deterministic contract (YAML) that includes: • name (derived from the spec)• mode: local• steps (checkout, lint, unit‑test, package‑audit, report)• artifacts (logs, test‑reports, package‑sbom) |
The contract is written to ./ricky-output/<workflow‑id>.yaml. |
| 4️⃣ | Skip execution – Because --run is omitted, the Execution‑Engine gate is not invoked. No containers are started, no checks are performed. |
You end up with a **static wo |
| ...[truncated] |
**Must**
- Say generation is the default and execution was not requested.
- Print the generated artifact path, workflow id, spec digest, and next run command.
- Avoid showing execution evidence for a generation-only request.
**Must Not**
- Imply the workflow ran automatically.
- Present a generation-only result as execution success.
- Hide the opt-in commands for running the artifact.
**Deterministic checks**
- PASS `status`: case executed
- PASS `maxToolCalls`: expected <= 0 tool calls, got 0
</details>
<details>
<summary><strong>REVIEW</strong> <code>cli.first-run-copy-is-compact-and-truthful</code> (cli-behavior/manual)</summary>
**User message**
> Render Ricky's first-run CLI onboarding for a new user.
**Ricky output**
```text
**Ricky – First‑Run CLI Onboarding**
Below is the complete, ready‑to‑run onboarding workflow for a brand‑new user of the **Ricky** AgentWorkforce CLI. It follows the repository conventions (AGENTS.md), the workflow standards, and the shared authoring rules. The contract is deterministic, includes verification gates, and produces the required review artifacts.
---
## 1️⃣ Workflow Contract
| Item | Description |
|------|-------------|
| **Name** | `ricky-onboard-first-run` |
| **Scope** | New‑user local environment setup (OS ≥ Ubuntu 20.04, macOS 12+, Windows 10+). |
| **Inputs** | None (interactive). |
| **Outputs** | • `~/.ricky/config.yaml` (validated config) <br>• `~/.ricky/credentials.json` (encrypted) <br>• `ricky version` printed <br>• Success flag file `~/.ricky/.onboard_complete` |
| **Pre‑conditions** | • `git` ≥ 2.30, `python` ≥ 3.9, `pip` ≥ 22.0 installed. <br>• Network connectivity to `pypi.org` and `github.com`. |
| **Post‑conditions** | • CLI commands (`ricky status`, `ricky workflow list`) run without error. <br>• No stray temporary files left in `/tmp`. |
| **Verification Gates** | 1. **Dependency Check** – abort if any required binary missing. <br>2. **Config Sc
...[truncated]
Must
- Show compact Ricky branding and clear Local / BYOH, Cloud, Both, and Just explore choices.
- End every branch with a concrete next step.
- Advertise only commands that are currently implemented.
Must Not
- Sound like a launch page or documentation dump.
- Claim Ricky runs workflows by default when generation is the default path.
- Require web or Slack onboarding before CLI use.
Deterministic checks
- PASS
status: case executed - PASS
maxToolCalls: expected <= 0 tool calls, got 0
REVIEW cli.recovery-guidance-no-stack-traces (cli-behavior/manual)
User message
A user gives Ricky an empty spec or a missing spec file.
Ricky output
**Ricky’s Standard Response to an Empty or Missing Spec File**
---
### 1. Immediate Detection & Blocker Report
| Step | Action | Output |
|------|--------|--------|
| **1.1** | Verify the presence of the spec file at the expected path (`/specs/<workflow>.yaml`). | If the file is not found, raise a **Blocker**: <br>`🚨 BLOCKER: Spec file “<workflow>.yaml” is missing.` |
| **1.2** | If the file exists but is empty (0 bytes) or contains only comments/whitespace, raise a **Blocker**: <br>`🚨 BLOCKER: Spec file “<workflow>.yaml” is empty or contains no actionable content.` |
| **1.3** | Log the blocker in the workflow’s **#issues** channel and tag the requestor. | `@requestor Please provide a complete spec or confirm the intended workflow name.` |
---
### 2. Create a Minimal Placeholder (Optional, for traceability)
If the requestor cannot immediately supply a spec, generate a **placeholder** that satisfies the repository’s contract and makes the missing‑spec state explicit.
```yaml
# /specs/<workflow>.yaml (placeholder)
# -------------------------------------------------
# NOTE: This is a generated placeholder. The real
# workflow definition is pending from the requestor.
# -----
...[truncated]
Must
- Return a user-facing failure or guidance message with a real recovery command.
- Distinguish generation failure from execution failure.
- Show stack traces only when verbose diagnostic mode is requested.
Must Not
- Crash with an uncaught exception in normal mode.
- Suggest commands that do not exist.
- Pretend a missing spec was accepted.
Deterministic checks
- PASS
status: case executed - PASS
maxToolCalls: expected <= 0 tool calls, got 0
REVIEW cli.status-does-not-invent-provider-state (cli-behavior/manual)
User message
Render
ricky statuswhen no provider checks have proven Google or GitHub are connected.
Ricky output
**`ricky status` output (no provider checks succeeded for Google or GitHub)**
=== AgentWorkforce Status Report ===
Timestamp: 2026‑05‑14T⟨HH:MM:SS⟩Z
🟠 Provider Connectivity
• Google : ❌ NOT CONNECTED (no successful health‑check)
• GitHub : ❌ NOT CONNECTED (no successful health‑check)
🔧 Workflow Engine
• Scheduler : ✅ RUNNING
• Queue : 0 pending jobs
• Workers : 4 active / 4 idle
📦 Repository State
• Branches : main (up‑to‑date)
• PR Queue : empty
• Last Commit : (⟨date⟩)
• Google OAuth token missing or invalid.
• GitHub personal‑access token missing or invalid.
• Dependent workflows that require Google or GitHub access are paused.
🛠️ Recommended Actions
- Verify OAuth credentials for Google and store them in the
GOOGLE_TOKENsecret. - Verify GitHub PAT and store it in the
GITHUB_TOKENsecret. - Re‑run provider health checks:
ricky provider check google ricky provider check github
- After successful checks, re‑enable any paused workflows:
ricky workflow resume <workflow‑name>
📚 Documentation
• Provider configuration: https
...[truncated]
**Must**
- Report unknown or not-connected provider state honestly.
- Update provider status only from explicit provider checks or Cloud status results.
- Give concrete setup guidance for Cloud when relevant.
**Must Not**
- Mark Google or GitHub connected because guidance text was shown.
- Invent a provider connection URL or OAuth flow.
- Show empty fields with no recovery guidance when config is missing.
**Deterministic checks**
- PASS `status`: case executed
- PASS `maxToolCalls`: expected <= 0 tool calls, got 0
</details>
<details>
<summary><strong>REVIEW</strong> <code>generation-quality.skill-matcher-registry-backed</code> (generation-quality/manual)</summary>
**User message**
> Generate a workflow for a GitHub primitive change and explain which skills Ricky selected.
**Ricky output**
```text
## GitHub Primitive‑Change Workflow
**Purpose:** Apply a single‑file edit (e.g., README typo fix) to a public repository while guaranteeing reproducible review, testing, and merge.
### 1. Workflow Contract (YAML‑style summary)
```yaml
name: primitive-change
description: One‑file edit → PR → CI → Merge
trigger:
- manual (Ricky UI) or API call
inputs:
repo: # full repo slug “org/project”
branch: main # target branch
file_path: # relative path to edited file
new_content: # full file contents (or diff patch)
commit_message: # PR title & body
reviewers: [] # GitHub usernames (optional)
outputs:
pr_url: string
merge_status: enum[merged, closed, blocked]
gates:
- syntax_check
- lint
- unit_test (if applicable)
- review_approval
- merge_protection
validation_loops: 80-100 # max attempts per gate before blocker
2. Step‑by‑Step Execution Plan
| # | Step | Owner | Artifact | Deterministic Gate |
|---|---|---|---|---|
| 1 | Initialize – Ricky clones repo at branch. |
Ricky (automation) | clone.log |
– |
| 2 | Apply Change – Write new_content to `file_pat |
|||
| ...[truncated] |
**Must**
- Select skills from the actual registry rather than a hardcoded static set.
- Record ranked skill matches with confidence and trigger evidence.
- Fall back gracefully when the registry is missing or no skill clears the threshold.
**Must Not**
- Claim runtime agents embody skills when only generation-time guidance was applied.
- Fail generation solely because optional skill files are missing.
- Hide skill selection evidence from artifacts or JSON output.
**Deterministic checks**
- PASS `status`: case executed
- PASS `maxToolCalls`: expected <= 0 tool calls, got 0
</details>
<details>
<summary><strong>REVIEW</strong> <code>generation-quality.tool-selector-honors-spec-hints</code> (generation-quality/manual)</summary>
**User message**
> Generate a workflow where the spec says "use Claude to review and Codex to implement".
**Ricky output**
```text
**Workflow Title:** Claude‑Review → Codex‑Implementation Pipeline
**Purpose**
Automate a deterministic hand‑off where Claude (LLM‑reviewer) validates design/spec changes and Codex (code‑generator) produces the implementation. The workflow enforces reproducible contracts, verification gates, and explicit review artifacts to guarantee reliability and traceability.
---
## 1. Contract Overview
| Element | Definition |
|---------|------------|
| **Input Artifact** | `design_spec.md` – Markdown spec located in `specs/` (UTF‑8, ≤ 200 KB). |
| **Reviewer** | Claude (model `claude-3-opus`) invoked via `ricky invoke claude`. |
| **Generator** | Codex (model `codex-12b`) invoked via `ricky invoke codex`. |
| **Output Artifacts** | 1. `review_report.json` (Claude) 2. `implementation/` (generated source files) 3. `ci_report.xml` (test results). |
| **Success Criteria** | • Claude rating ≥ 4/5 on “spec completeness”. <br>• All generated unit tests pass (≥ 90 % coverage). <br>• No blocker flagged in `review_report.json`. |
| **Failure Modes** | • Rating < 4 → abort & raise blocker. <br>• Test failures → auto‑re‑run Codex up to 3 attempts, then raise blocker. |
| **Versioning** | Branch
...[truncated]
Must
- Assign per-step runner/model decisions from explicit spec hints where possible.
- Let skill metadata or project defaults fill gaps when the spec is silent.
- Write
tool-selection.jsonor equivalent audit metadata.
Must Not
- Use one generic runtime default for every agent despite explicit hints.
- Let skill defaults override explicit user/spec runner hints.
- Omit the reason each tool or model was chosen.
Deterministic checks
- PASS
status: case executed - PASS
maxToolCalls: expected <= 0 tool calls, got 0
REVIEW generation-quality.refine-is-opt-in-and-bounded (generation-quality/manual)
User message
Use
--refineto sharpen a generated workflow's step task descriptions and acceptance gates.
Ricky output
### Using `--refine` to Sharpen Step Descriptions & Acceptance Gates
`--refine` is a deterministic post‑processing flag that takes a **generated workflow JSON** (or YAML) and returns a **tighter, review‑ready version**. It focuses on two things:
| Target | What `--refine` does |
|--------|----------------------|
| **Task descriptions** | • Removes vague verbs (“do”, “handle”). <br>• Adds concrete inputs, outputs, and success criteria. <br>• Enforces the *“who‑does‑what‑when‑how”* template. |
| **Acceptance gates** | • Converts informal “check” statements into explicit **verification predicates** (e.g., `output.fileExists && output.size > 0`). <br>• Inserts **deterministic verification steps** (unit test, schema validation, checksum). <br>• Adds a **gate‑type tag** (`gate: unit-test`, `gate: schema`, `gate: smoke`). |
---
## 1. Basic Invocation
```bash
agentworkforce generate workflow.yaml --model gpt‑4o-mini > raw_workflow.yaml
agentworkforce refine raw_workflow.yaml --out refined_workflow.yaml
raw_workflow.yaml– the initial draft (may contain free‑form prose).refined_workflow.yaml– the output with sharpened tasks and explicit gates.
You can also pipe di
...[truncated]
**Must**
- Keep the deterministic unrefined path as the default.
- Bound the refinement pass by timeout, token budget, and editable regions.
- Re-run validation after refinement and fall back to the deterministic artifact on unsafe edits or provider failure.
**Must Not**
- Change the workflow graph, agent assignments, or side-effect scope during refinement.
- Fail the whole generation if optional refinement times out.
- Hide warnings when the deterministic artifact is returned unchanged.
**Deterministic checks**
- PASS `status`: case executed
- PASS `maxToolCalls`: expected <= 0 tool calls, got 0
</details>
<details>
<summary><strong>REVIEW</strong> <code>generation-quality.behavior-grounded-gates</code> (generation-quality/manual)</summary>
**User message**
> Generate a workflow for the `ricky --version` spec.
**Ricky output**
```text
**Workflow: `ricky --version` Specification & Delivery**
*(Ricky AgentWorkforce – v0.9.3)*
---
### 1. Objective
Produce a deterministic, test‑driven implementation of the `ricky --version` CLI flag that:
* Returns the current Ricky version string (e.g., `ricky 0.9.3`).
* Exits with status 0 on success, non‑zero on unexpected failure.
* Is fully covered by unit tests, integration tests, and CI verification gates.
* Is delivered via a scoped branch/PR with all required review artifacts.
---
### 2. Contract (Interface Definition)
| Element | Description |
|---------|-------------|
| **Command** | `ricky --version` |
| **Output** | `ricky <semver>` printed to **stdout** (no trailing newline required). |
| **Exit Code** | `0` on success, `>0` on any internal error (e.g., version file missing). |
| **Side‑effects** | None (pure read‑only). |
| **Supported Platforms** | Linux, macOS, Windows (via the same entry point). |
| **Version Source** | Single source of truth: `src/__init__.py` variable `__version__ = "0.9.3"` (or the value from `pyproject.toml` under `[tool.poetry.version]`). |
| **Error Message** | If version cannot be resolved, print `Error: unable to determine Ric
...[truncated]
Must
- Build gates from the stated acceptance behavior, such as checking
ricky --versionoutput. - Avoid generic source-shape grep checks when the spec asks for CLI behavior.
- Keep generated validation meaningful for the current repo shape.
Must Not
- Treat
grep -Eq 'export|function|class|workflow(' dist/ricky.jsas proof of version behavior. - Claim the workflow is proven by source syntax alone.
- Ignore the package-json version resolution order in the spec.
Deterministic checks
- PASS
status: case executed - PASS
maxToolCalls: expected <= 0 tool calls, got 0
REVIEW generation-quality.pattern-selection-deliberate (generation-quality/manual)
User message
Generate a workflow for many independent artifacts with a validation/fix/rerun loop.
Ricky output
**Workflow: “Batch‑Artifact Validation‑Fix‑Rerun Loop”**
*Purpose*: Process a large collection of independent artifacts (e.g., data files, code modules, model checkpoints) through a deterministic validation pipeline, automatically apply fix‑scripts where possible, and rerun until all artifacts are either **PASS** or **BLOCKED**. The workflow is scoped to a single feature branch (`batch‑validation‑loop`) and produces a single PR that contains only the generated CI configuration and supporting scripts.
---
## 1. Repository Layout (branch `batch‑validation‑loop`)
/.github/workflows/
batch_validation.yml # CI definition (this workflow)
/scripts/
validate.py # deterministic validator (returns JSON)
fix.py # idempotent fixer (optional per‑artifact)
report.py # aggregates results, creates PR comment
/config/
validation_schema.json # JSON‑Schema for artifact validation
/artifacts/
*.json / *.csv / *.py … # target artifacts (any type, independent)
All new files are added under `/scripts/` and `/config/`. No existing production code is modified.
---
## 2. Contract (Inputs / O
...[truncated]
Must
- Choose
dag,supervisor, orpipelinedeliberately based on the work shape. - Explain the pattern choice in artifact metadata or a rationale.
- Use
dagfor validation/fix/rerun loops when dependencies matter.
Must Not
- Default blindly to
dagfor every workflow. - Collapse independent artifact work into one vague agent task.
- Omit verification gates because the chosen pattern seems obvious.
Deterministic checks
- PASS
status: case executed - PASS
maxToolCalls: expected <= 0 tool calls, got 0
REVIEW generation-quality.skill-boundary-copy (generation-quality/manual)
User message
Describe how selected workflow-writing skills affected a generated Ricky workflow.
Ricky output
### How Selected Workflow‑Writing Skills Shaped the Generated Ricky Workflow
| Skill | What the Skill Enforces | Direct Effect on the Generated Workflow |
|-------|------------------------|------------------------------------------|
| **Concrete Contract Definition** | Every step must declare its **input contract** (type, required fields, validation rules) and **output contract** (shape, optional fields, version). | The workflow begins with a `contract` block for the trigger event (`event.type: “ticket.created”`). Down‑stream steps reference `{{stepX.output}}` only after the contract is satisfied, preventing accidental field leakage. |
| **Deterministic Verification Gates** | After each logical block, a **gate** runs a pure‑function test that either passes or throws a deterministic error. | After the “Enrich Ticket” step, a gate checks `output.priority ∈ {P1,P2,P3,P4}`. If the gate fails, the workflow aborts with a clear `GateFailure: priority‑validation`. This makes the path reproducible and debuggable. |
| **Review Artifacts Generation** | The author must emit a **review‑artifact** markdown file that lists step purpose, contract diff, and test coverage. | The generated workflo
...[truncated]
Must
- Describe skills as generation-time selection, loading, and rendering inputs.
- Point to metadata such as
loaded-skills.txtandskill-application-boundary.json. - Say the workflow instructions were informed by selected skills.
Must Not
- Say runtime agents are skill-embedded, skill-powered at runtime, or embody skills unless runtime skill loading is implemented and tested.
- Treat metadata existence as proof that agents received skill bodies at runtime.
- Overstate current tests beyond the generation boundary.
Deterministic checks
- PASS
status: case executed - PASS
maxToolCalls: expected <= 0 tool calls, got 0
REVIEW generation-quality.no-pure-codegen-without-proof (generation-quality/manual)
User message
Generate a workflow from a vague product spec and return it to the user.
Ricky output
**Ricky – Workflow Generation Assistant**
---
### 📋 Quick Summary
To turn a product specification into a reliable, review‑ready workflow we need a **clear, concrete contract** that defines:
1. **Inputs** – data types, sources, and validation rules.
2. **Processing steps** – deterministic actions, error‑handling, and any external calls.
3. **Outputs** – artifacts, formats, and downstream hand‑off points.
4. **Verification gates** – unit‑test suites, integration checks, and manual‑review checkpoints.
5. **Review artifacts** – design doc, test plan, and PR checklist.
Because the spec you provided is vague, I’ll ask for the missing details needed to create a scoped, deterministic workflow that meets our **80‑to‑100 validation loop** standard.
---
## 🔎 Information Needed
| Area | What I Need from You | Why It Matters |
|------|----------------------|----------------|
| **Goal & Scope** | A one‑sentence description of the product’s primary function (e.g., “Generate a PDF invoice from an order JSON”). | Defines the end‑to‑end boundary of the workflow. |
| **Inputs** | • Data format (JSON, CSV, etc.) <br>• Source (API, file upload, DB) <br>• Required fields & validat
...[truncated]
Must
- Produce a Relay-native TypeScript workflow with explicit verification, review, and signoff.
- Validate with dry-run or targeted structural checks where possible.
- Return artifacts, warnings, and follow-up commands honestly.
Must Not
- Act like Ricky is a pure code-generation bot that emits workflows without verification.
- Stop at "code compiles" as the proof bar.
- Skip skill-aware workflow authoring guidance for serious workflows.
Deterministic checks
- PASS
status: case executed - PASS
maxToolCalls: expected <= 0 tool calls, got 0
REVIEW runtime-recovery.classify-before-retry (runtime-recovery/manual)
User message
A workflow failed after one step timed out and another worker stayed alive without producing artifacts. Explain what Ricky should do next.
Ricky output
**Ricky’s post‑failure protocol**
1. **Detect & Log the anomaly**
- The orchestrator reports *Step X* timed‑out (exceeded its SLA).
- *Worker Y* is still “alive” (heartbeat present) but has not emitted any artifact or status update.
2. **Mark the step as failed**
- Update the workflow state machine: `StepX.status = FAILED_TIMEOUT`.
- Record the exact timestamp, configured timeout value, and any partial logs that were captured.
3. **Isolate the stray worker**
- Send a *graceful‑shutdown* command to Worker Y.
- If the worker does not acknowledge within a short grace period (e.g., 30 s), issue a *hard‑kill* (container stop / process kill).
- Record the shutdown outcome in the audit log.
4. **Generate a deterministic verification gate**
- **Gate G‑timeout‑artifact**:
- Preconditions: `StepX.status == FAILED_TIMEOUT` **AND** `WorkerY.state == TERMINATED`.
- Postconditions: No artifact for Step X exists **AND** a *failure ticket* is created.
- The gate must pass before any downstream steps are allowed to proceed.
5. **Create the failure ticket (review artifact)**
- Title: *Workflow <WF‑ID> – Step X timeout & orphan work
...[truncated]
Must
- Classify the failure before deciding whether to retry.
- Distinguish agent-runtime opacity, timeout, environment blockers, workflow-structure bugs, and validation-strategy mismatch.
- Preserve raw evidence and uncertainty when the class is not known.
Must Not
- Blindly rerun the whole workflow without a blocker classification.
- Treat every failure as a broken workflow definition.
- Claim Ricky fixed the workflow before rerun evidence proves it.
Deterministic checks
- PASS
status: case executed - PASS
maxToolCalls: expected <= 0 tool calls, got 0
REVIEW runtime-recovery.stale-relay-state (runtime-recovery/manual)
User message
Ricky detects stale
.agent-relay/,.relay/, and.trajectories/state before launching a local workflow.
Ricky output
**Workflow Contract – Stale‑State Detection for Local Launches**
| **Element** | **Definition** |
|-------------|----------------|
| **Name** | `detect‑stale‑state` |
| **Scope** | Runs **once**, immediately before any local workflow is started (e.g., `ricky run …`). |
| **Inputs** | - Working directory (root of the repo).<br>- Optional config `ricky.yaml` flag `staleCheck.enabled` (default = true). |
| **Outputs** | - `stale: true|false` (boolean).<br>- `stalePaths: []string` (list of stale directories/files).<br>- `report: string` (human‑readable summary). |
| **Side‑effects** | - If `stale == true`, the workflow **fails fast** with a deterministic error code `ERR_STALE_STATE`.<br>- Writes a JSON artifact `./.ricky/artifacts/stale‑state.json`. |
| **Deterministic Verification Gate** | `verify‑stale‑state` (see below). |
| **Failure Mode** | Blocker – any stale state must be resolved before proceeding. |
---
### 1. Detection Logic (Implementation Sketch)
```yaml
# .ricky/workflows/detect-stale-state.yaml
name: detect-stale-state
description: Detect stale .agent‑relay/, .relay/, and .trajectories/ state before a local run.
runs:
using: composite
steps:
- id: scan
...[truncated]
Must
- Classify stale local runtime state as an environment contamination issue.
- Recommend quarantine or isolated-run guidance before launch.
- Record the observed paths and the action taken or recommended.
Must Not
- Treat stale runtime state as a workflow logic failure.
- Delete or overwrite state without an explicit safe path or user intent.
- Continue into execution as if the workspace were clean.
Deterministic checks
- PASS
status: case executed - PASS
maxToolCalls: expected <= 0 tool calls, got 0
REVIEW runtime-recovery.already-running-conflict (runtime-recovery/manual)
User message
A run marker says another Ricky or Relay run is already active in this workspace.
Ricky output
**How to clear the “another Ricky/Relay run is already active in this workspace” marker**
The marker is a lock file that prevents two concurrent executions of the same workspace. It can be left behind if a previous run crashed, was killed, or if you accidentally started a second instance. Follow the deterministic clean‑up contract below.
---
## 1. Verify the current lock state (read‑only)
```bash
# From the root of the workspace
ricky status # shows active runs, PID, and lock file path
cat .ricky/run.lock # optional: view raw lock contents
You should see something like:
Lock file: .ricky/run.lock
Owner PID: 12345
Started: 2026‑05‑14T09:32:11Z
Command: ricky run …
If the PID does not exist (e.g., ps -p 12345 returns nothing) the lock is stale.
2. Deterministic clean‑up gate
| Condition | Action |
|---|---|
| PID is alive and you intentionally started another run | Abort the new run. Wait for the first run to finish or cancel it with ricky cancel <run‑id>. |
| PID is alive but the process is hung or you need to pre‑empt it | Terminate the process (kill -9 12345) only after confirming no other team member |
| ...[truncated] |
**Must**
- Report the active marker, run id, or status path when available.
- Ask the user to inspect, wait for, or explicitly clear the active run.
- Avoid launching a competing run that could corrupt evidence.
**Must Not**
- Silently start another run.
- Hide the existing run marker from the user.
- Treat the conflict as a generic failure with no recovery path.
**Deterministic checks**
- PASS `status`: case executed
- PASS `maxToolCalls`: expected <= 0 tool calls, got 0
</details>
<details>
<summary><strong>REVIEW</strong> <code>runtime-recovery.env-loader-injection-runtime-loadable</code> (runtime-recovery/manual)</summary>
**User message**
> A workflow artifact references a `MISSING_ENV_VAR` value. Ricky's deterministic auto-fix injects the `.env.local` / `.env` loader (`loadRickyWorkflowEnv`) and the optional `assertRickyWorkflowEnv` guard into the artifact before retry. The artifact may be a master-rendered workflow whose `.step({ command: ... })` bodies embed `node --input-type=module` HEREDOCs containing literal `import { ... } from 'node:fs'` / `from 'node:path'` strings.
**Ricky output**
```text
**Deterministic auto‑fix for a workflow that references a missing environment variable**
Below is the concrete, repeatable contract you can apply to any workflow artifact that fails because it tries to read `process.env.MISSING_ENV_VAR` (or any other undefined variable). The fix injects the Ricky environment loader and optional guard, then re‑renders the step bodies that contain inline `node --input-type=module` HEREDOCs.
---
## 1. Preconditions (Verification Gate)
| Check | Expected | Action if not met |
|------|----------|-------------------|
| Artifact is a **master‑rendered** workflow (i.e. a `.js` file that exports a `workflow` object). | ✅ | Abort – not a supported artifact. |
| The artifact contains at least one `process.env.MISSING_ENV_VAR` reference (or any `process.env.*` that is undefined at load time). | ✅ | Continue. |
| The artifact uses **step‑level HEREDOCs** with `node --input-type=module` (e.g. ```.step({ command: `node --input-type=module <<'EOF'\nimport …\nEOF` })```). | ✅ | Continue. |
| A `.env` or `.env.local` file exists at the repository root (or can be created). | ✅/❌ | If missing, create an empty `.env.local` file (git‑ignored). |
If any gate fails,
...[truncated]
Must
- Produce a repaired artifact that successfully loads under Node, not just one that contains the marker comment. The injected
loadRickyWorkflowEnvbody referencesrickyWorkflowFs.*andrickyWorkflowPath.*, so the repair must also add the correspondingimport * as rickyWorkflowFs from 'node:fs'andimport * as rickyWorkflowPath from 'node:path'aliases at module top level. - Detect existing alias imports by matching real top-level
import * as <alias> from '<module>'statements, not by substring-matching the module specifier anywhere in the file (substrings inside HEREDOCs in.step({ command: ... })bodies do not count as imports). - Leave the embedded shell HEREDOC contents untouched so the runtime-spawned child process still sees the literal import lines it expects.
Must Not
- Skip adding the
rickyWorkflowFs/rickyWorkflowPathaliases becausefrom 'node:fs'orfrom 'node:path'already appears somewhere in the file as a string literal. - Inject
loadRickyWorkflowEnv(orassertRickyWorkflowEnv) without the supporting alias imports, which produces aReferenceError: rickyWorkflowPath is not definedat module load and burns the auto-fix budget onUNSUPPORTED_RUNTIME at runtime-launch. - Rewrite or escape the embedded HEREDOC text in step commands.
Deterministic checks
- PASS
status: case executed - PASS
maxToolCalls: expected <= 0 tool calls, got 0
REVIEW runtime-recovery.auto-fix-bounded-loop (runtime-recovery/manual)
User message
Run a local workflow with auto-fix enabled. The first attempt fails, the workflow artifact is repairable, and the failed step plus previous run id are available.
Ricky output
**Local Workflow Execution – Auto‑Fix Mode**
Below is a deterministic, step‑by‑step contract you can copy‑paste into your terminal (or CI script) to run the workflow, detect the failure, invoke the auto‑fix routine, and re‑run the repaired artifact. All verification gates and review artifacts are included so you can audit each loop.
---
## 1️⃣ Prerequisites (one‑time)
| Item | Check | Command |
|------|-------|---------|
| `ricky` CLI installed (≥ 2.4) | `ricky --version` | `pip install ricky-workflow` |
| Workflow repo cloned & on a clean branch | `git status` shows “nothing to commit” | `git checkout -b wf‑auto‑fix‑run` |
| Docker (or the runtime specified in `workflow.yaml`) | `docker version` | Install Docker CE if missing |
| Access to the previous run ID (e.g., `run‑20240512‑a1b2c3`) | Provided by user | – |
| Failed step name (e.g., `build‑docker‑image`) | Provided by user | – |
---
## 2️⃣ Input Contract (what you must supply)
```json
{
"workflow_id": "my_project/pipelines/ci.yml", // path to the workflow definition
"run_id": "run-20240512-a1b2c3", // previous successful run (used for cache)
"failed_step": "build-docker-image", // ex
...[truncated]
Must
- Use a bounded retry budget and summarize every attempt.
- Ask the Workforce workflow persona to repair the workflow artifact when a resolvable artifact exists.
- Resume from the failed step with the previous run id when those values are available.
Must Not
- Edit arbitrary repository source files as the default auto-fix surface.
- Keep retrying after the configured max attempts.
- Lose the single Ricky tracking run id across repair/resume attempts.
Deterministic checks
- PASS
status: case executed - PASS
maxToolCalls: expected <= 0 tool calls, got 0
REVIEW runtime-recovery.no-auto-fix-preserves-single-attempt (runtime-recovery/manual)
User message
A user runs
ricky run workflows/foo.ts --no-auto-fixand the workflow fails.
Ricky output
**Quick‑Start Failure Checklist**
| Step | Action | Why it matters |
|------|--------|----------------|
| 1️⃣ | **Inspect the CLI output** – look for the *error code* and the *stack trace* printed just before the process exits. | The first line usually tells you which validation gate failed (e.g., `ContractViolation`, `TypeCheck`, `RuntimeAssert`). |
| 2️⃣ | **Open the workflow log** – `ricky logs workflows/foo.ts` (or check `~/.ricky/logs/foo-<timestamp>.json`). | Logs contain the deterministic verification gate results, the exact input payload, and any side‑effect snapshots. |
| 3️⃣ | **Run in debug mode** – `ricky run workflows/foo.ts --debug --no-auto-fix`. | Adds step‑by‑step tracing, prints the intermediate contract payloads, and disables any silent retries. |
| 4️⃣ | **Validate the contract file** – open `workflows/foo.contract.ts` (or `.json`) and run `ricky contract validate workflows/foo.contract.ts`. | Guarantees the contract is syntactically correct and that all required fields are present. |
| 5️⃣ | **Check the branch/PR boundary** – ensure you are on the branch that contains the latest PR for `foo.ts`. If the workflow was edited in a PR that isn’t merged, the runtime
...[truncated]
Must
- Preserve one-attempt behavior when auto-fix is disabled.
- Return the classified blocker, diagnosis, recovery steps, and non-zero exit code.
- Make clear that the user chose manual inspection over repair/resume automation.
Must Not
- Start a repair loop despite
--no-auto-fix. - Suppress the diagnosis because no repair was attempted.
- Present the failure as a completed repair attempt.
Deterministic checks
- PASS
status: case executed - PASS
maxToolCalls: expected <= 0 tool calls, got 0
REVIEW runtime-recovery.in-process-local-runner (runtime-recovery/manual)
User message
Explain how Ricky should execute a local TypeScript workflow artifact in the primary local path.
Ricky output
**Ricky’s Standard Procedure for Executing a Local TypeScript Workflow Artifact**
Below is the concrete, repeatable contract that Ricky follows whenever a developer asks to run a TypeScript‑based workflow that lives in the *primary local path* (by convention `./workflows`). The steps are written as a deterministic “execution gate” that can be scripted, reviewed, and automatically verified.
---
## 1. Preconditions (Gate A – Environment Validation)
| Check | Command | Expected Result | Failure Action |
|-------|---------|----------------|----------------|
| Node version ≥ 18 | `node -v` | `v18.x` or higher | Abort – report “Node version too low”. |
| npm ≥ 9 | `npm -v` | `9.x` or higher | Abort – report “npm version too low”. |
| `ts-node` installed globally or locally | `npx ts-node -v` | `ts-node vX.Y.Z` | If missing, run **Gate B** to install. |
| Workspace clean (no uncommitted changes) | `git status --porcelain` | empty output | Abort – ask user to stash/commit. |
| `package.json` present in `./workflows` | `test -f ./workflows/package.json` | file exists | Abort – ask user to create a minimal `package.json`. |
If **all** checks pass, proceed to Gate B. If any check fails,
...[truncated]
Must
- Prefer the Node strip-types route or equivalent SDK/programmatic route over requiring the
agent-relaybinary on PATH. - Precheck that Node and
@agent-relay/sdkare resolvable for the workflow. - Record the actual spawn command in execution evidence.
Must Not
- Fail solely because
agent-relayis not on PATH when the SDK route is available. - Hide the actual runtime command from evidence.
- Conflate the user-facing reproduction command with the primary internal spawn route.
Deterministic checks
- PASS
status: case executed - PASS
maxToolCalls: expected <= 0 tool calls, got 0
REVIEW runtime-recovery.escalation-is-not-generic-failure (runtime-recovery/manual)
User message
Ricky reaches a boundary after a structural failure persists after a fix attempt.
Ricky output
**Workflow: “Structural‑Failure Escalation – Persistent Failure after Fix Attempt”**
*(Ricky – AgentWorkforce reliability & coordination)*
---
## 1. Scope & Ownership
| Element | Detail |
|---|---|
| **Trigger** | A structural failure (e.g., service outage, data‑pipeline break, infra‑component collapse) has been **re‑opened** after an initial remediation was merged. |
| **Owner** | **Reliability Engineer (RE)** – primary executor. <br>**Escalation Lead (EL)** – reviewer & gatekeeper. |
| **Stakeholders** | SRE team, Product Owner, Incident Commander, QA, Architecture Review Board (ARB). |
| **Branch** | `fix/<service>-<issue-id>` → `escalate/persistent‑<issue-id>` (protected). |
| **PR Target** | `main` (or the service‑specific release branch). |
---
## 2. Contract (Inputs / Outputs)
| Contract Item | Description |
|---|---|
| **Input Artifact** | `failure_report_<id>.json` – includes: <br>• Original failure ID, timestamps, logs, metrics. <br>• Fix commit SHA, diff, and deployment version. <br>• Re‑open reason (e.g., regression test failure, post‑deploy alert). |
| **Output Artifact** | `escalation_bundle_<id>.zip` – contains: <br>• Updated root‑cause analysis (RCA) markdo
...[truncated]
Must
- Escalate with the attempted fix, failed validation, classified blocker, and recommendation.
- Distinguish escalation from a generic product failure.
- Preserve enough context for a human operator to continue.
Must Not
- Retry speculative fixes indefinitely.
- Collapse to "something went wrong" without the attempted actions.
- Discard evidence from failed repair attempts.
Deterministic checks
- PASS
status: case executed - PASS
maxToolCalls: expected <= 0 tool calls, got 0
REVIEW runtime-recovery.analytics-from-structured-evidence (runtime-recovery/manual)
User message
Produce a workflow health digest from many Ricky workflow runs.
Ricky output
**Ricky Workflow – Health‑Digest Generator**
*Purpose:* Aggregate the results of multiple Ricky workflow runs, compute key health metrics, and emit a concise “Health Digest” report (Markdown) for distribution to the Ops team.
---
## 1. Contract Overview
| Item | Specification |
|------|----------------|
| **Input** | `run‑metadata.jsonl` – newline‑delimited JSON objects, one per workflow run. Each object must contain the fields listed in **Section 2**. |
| **Output** | `health‑digest.md` – Markdown file containing a summary table, trend charts (as embedded Base64 PNGs), and a “Blockers” section. |
| **Success Criteria** | - All required input fields are present for ≥ 95 % of runs.<br>- No validation errors are emitted.<br>- Digest file is written and passes schema validation (see § 4). |
| **Failure Modes** | - Missing required fields → **BLOCKER** (abort, raise issue).<br>- JSON parse error → **BLOCKER** (abort).<br>- No runs in the time window → **WARN** (empty digest). |
| **Idempotency** | The workflow is pure: given the same `run‑metadata.jsonl` it always produces the same `health‑digest.md`. |
| **Deterministic Gates** | 1️⃣ Input validation<br>2️⃣ Aggregation sanity c
...[truncated]
Must
- Consume normalized
WorkflowRunEvidencerather than raw logs as the primary input. - Identify recurring failure classes, weak validation, oversized steps, and runtime duration patterns.
- Produce concrete recommendations tied to specific workflows, steps, or metrics.
Must Not
- Mutate evidence while analyzing it.
- Return generic advice like "improve your workflow" without references.
- Mix environment failures and workflow-logic failures into one undifferentiated bucket.
Deterministic checks
- PASS
status: case executed - PASS
maxToolCalls: expected <= 0 tool calls, got 0
REVIEW surfaces-ingress.co-equal-surfaces (surfaces-ingress/manual)
User message
Design a new Ricky Slack surface for workflow debugging.
Ricky output
**Ricky Slack Surface – “Workflow Debugger”**
*Version 0.1 – Design Draft*
---
## 1. Overview
The **Workflow Debugger** Slack surface gives engineers and product owners a deterministic, low‑latency UI for inspecting, stepping through, and fixing Ricky‑managed workflows directly from a Slack channel or DM. It follows the **Ricky repository conventions** (AGENTS.md) and the **workflow standards** (contract‑first, verification gates, review artifacts, 80‑100 validation loops, blocker reporting, scoped PR boundaries).
---
## 2. Interaction Model
| Slack Command | Trigger | Payload | Response | Follow‑up |
|---------------|---------|---------|----------|----------|
| `/ricky debug <workflow‑id>` | Slash command in any channel/DM | `{workflowId: string}` | Interactive message with **Workflow Summary Card** (status, last run, owner, version) and **Action Buttons** | `View Run`, `Step`, `Patch`, `Abort` |
| `View Run` | Button | `{runId: string}` | Modal showing **Run Timeline** (nodes, timestamps, inputs/outputs, logs) | Click node to open **Node Detail** |
| `Step` | Button (enabled only on paused runs) | `{runId, nodeId}` | Modal with **Step Controls** (Continue, Skip, Re‑run,
...[truncated]
Must
- Treat Slack as a co-equal product surface, not a wrapper around CLI.
- Route domain work through the same normalization, executor, evidence, and specialist stages.
- Keep Slack-specific formatting, thread handling, and interactive components in the surface layer.
Must Not
- Put workflow generation or diagnosis domain logic directly in the Slack handler.
- Degrade Slack to a developer shortcut with weaker routing than CLI.
- Skip signature verification, dedup, or thread handling when the surface is implemented.
Deterministic checks
- PASS
status: case executed - PASS
maxToolCalls: expected <= 0 tool calls, got 0
REVIEW surfaces-ingress.normalizer-is-only-translation-boundary (surfaces-ingress/manual)
User message
Add a new web handoff type that submits a workflow spec and mode.
Ricky output
**Feature:** Add a new *Web Handoff* type – **`WebSubmitSpec`** – that allows a client‑side form to submit a complete workflow specification together with an execution mode (e.g., `dry‑run`, `live`).
The change is scoped to the **`workflow‑handoff`** module and will be delivered in a dedicated branch **`feat/web‑handoff‑submit‑spec`** with a single PR.
---
## 1. Contract Definition (API)
| Field | Type | Required | Description |
|-------|------|----------|-------------|
| `spec` | `object` (JSON‑serializable) | ✅ | Full workflow spec as defined in `WorkflowSpec` (tasks, dependencies, contracts, etc.). |
| `mode` | `enum` {`dry-run`, `live`} | ✅ | Execution mode. `dry-run` validates and sim‑executes without side‑effects; `live` enqueues the workflow for real execution. |
| `metadata` | `object` (optional) | ❌ | Arbitrary key‑value pairs (e.g., `requestId`, `origin`). |
| `authToken` | `string` | ✅ | Bearer token for authorisation (validated by `AuthMiddleware`). |
**Endpoint**: `POST /api/v1/web-handoff/submit-spec`
**Content‑Type**: `application/json`
**Response**:
```json
{
"status": "accepted|rejected",
"workflowId": "<uuid>", // present if accepted
"
...[truncated]
Must
- Add a handoff variant that normalizes into
LocalInvocationRequestorCloudGenerateRequest. - Keep auth/session validation in the web surface before normalization.
- Ensure downstream code does not need to know the request came from web.
Must Not
- Short-circuit directly from the web handler to executors.
- Create a parallel domain model for web requests.
- Make the web surface the owner of local or Cloud routing semantics.
Deterministic checks
- PASS
status: case executed - PASS
maxToolCalls: expected <= 0 tool calls, got 0
REVIEW surfaces-ingress.mcp-claude-context-is-metadata (surfaces-ingress/manual)
User message
A Claude session hands Ricky a workflow spec plus conversation context and asks Ricky to determine whether to generate, debug, restart, analyze, or coordina
Comment truncated to stay within GitHub limits. Download the ricky-eval-run artifact for the full human-review.md.
Summary
Validation