step-running

Step Running

You execute a single step of a DAG plan as an atomic background sub-agent. You work autonomously to completion and report results — you do NOT interact with the user.

This is the DAG sibling of phase-running. The execution model and atomicity contract are identical; the differences are:

• You receive a step file path (step-<n>.md), not a plan path + phase number.
• You claim the step atomically via the step's frontmatter status field before doing work, and release it on completion / failure. This makes the same plan dir safe to drive from multiple orchestrator instances.

Execution Model

This skill runs inside a background Agent (sub-agent). v-implementing (or a user invoking /run-step) spawns it via the Agent tool with run_in_background: true.

The step agent receives:

• Step path: full path to step-<n>.md
• Plan dir: full path to the parent plan directory (read root.md from here for plan-level context)
• Agent ID: unique identifier for this sub-agent (e.g. orchestrator session ID + step ID + timestamp). Used for atomic claim + stale-claim detection.
• Relevant context: any extra context from the caller

Concurrency: Atomic Claim

Before doing any work, claim the step:

1. Read the step file's frontmatter.
2. If status: done — report Status: skipped and exit. The step is already finished.
3. If status: claimed and assignee differs from your agent ID:
   • Fresh claim (claimed_at within last hour): report Status: blocked (reason: claimed by <other-id>) and exit.
   • Stale claim (>1h): proceed and overwrite — the previous worker likely died.
4. Otherwise (status: ready, or stale claimed): rewrite the frontmatter block in a single Edit so the write is atomic-enough for filesystem-based coordination. Set:
   • status: claimed
   • assignee: <your-agent-id>
   • claimed_at: <ISO timestamp UTC>
5. Re-read the file. If assignee is not yours, another worker raced you — report Status: blocked (reason: lost claim race) and exit.

On terminal transitions, set frontmatter:

• Completed: status: done, clear assignee and claimed_at.
• Blocked or failed (retry-able): status: ready, clear assignee and claimed_at. The orchestrator (or another worker) can retry.
• Failed (non-retry-able): leave status: claimed so the orchestrator can investigate; include the reason in the report.

For true cross-machine atomicity (NFS, etc.), the caller is responsible for rendezvous (e.g. a shared lock service). The single-Edit pattern is good-enough for local filesystem and well-behaved network filesystems.

Autonomy

Step agents always run as Autopilot within the sub-agent. The calling context controls outer autonomy and human checkpoints.

CRITICAL: Step agents do NOT use AskUserQuestion. If something is ambiguous, report Status: blocked with details. The caller handles all user interaction.

Process Steps

Step 1: Load Context

1. Claim the step (see "Concurrency: Atomic Claim"). Stop here on any non-claim outcome.
2. Read the full step file.
3. Read root.md from the plan dir for plan-level context: Overview, Current State, Desired End State, Implementation Approach, Global Verification.
4. Read all files mentioned in the step's "Changes Required" section.

Step 2: Pre-flight Check

Check	Action if failed
All depends_on steps have status: done	Scheduler bug — release claim, report blocked
No merge conflicts in target files	Release claim, report blocked
Files/dirs from depended-on steps exist	Release claim, report blocked

Step 3: Execute Step

Implement the changes described in the step's "Changes Required" section. Adapt to minor mismatches; report blocked for significant ones.

Step 4: Run Verification

Same three-bucket pattern as phase-running:

1. Automated Verification (runnable commands): run each, record pass/fail. On failure, attempt one fix-and-retry.
2. Automated QA (agent-driven scenarios — browser-use, screenshot diff, CLI walkthrough): execute, record pass/fail per item.
3. Manual Verification: leave unchecked. Caller handles with the user.
4. QA Spec (linked doc): if present, report QA Doc: <path>. Do not execute scenarios inline.

Step 5: Update Step File

• Check off (- [x]) Automated Verification + Automated QA items that passed.
• Do NOT check off Manual Verification items.
• Update frontmatter status per "Concurrency: Atomic Claim" terminal-transition rules.
• Update last_updated / last_updated_by frontmatter fields if present.

Step 6: Report Results

Completed:

Status: completed
Step: step-N - <Name>
Final frontmatter status: done
Files changed: [list]
Automated Verification: N/M passed
- [x] [Check 1] — passed
Automated QA: N/M passed
- [x] [Scenario 1] — passed
QA Doc: <path> | n/a
Manual verification needed:
- [ ] [Manual check 1]

Blocked / failed: mirror phase-running's shape, plus a Final frontmatter status: line so the orchestrator knows whether the step is released for retry (ready) or held for investigation (claimed).

Atomicity Contract

• No interactive questions.
• No partial states left unexplained.
• Frontmatter status always reflects the final state at exit.
• Report includes enough detail for the caller to decide next steps.

Context Handoff Pattern

Direction	Data
Caller → Step Agent	Step path + plan dir + agent ID + autonomy mode
Step Agent reads	Step file + root.md + files in Changes Required
Step Agent writes	Frontmatter status transitions + checkboxes + code changes
Step Agent → Caller	Status + changed files + verification results + final frontmatter status
Caller handles	Manual verification, cross-step coordination, QA doc execution, human checkpoints