---
requirements:
  reasoning: Medium
  context: Medium
  speed: Medium
audience: subagent
phase: implement
context:
  architecture: false
  prior_summaries: delta
  persona: summary
  master_index: false
  diff_mode: false
deps:
  personas: [engineer]
  skills: [engineer, generic]
  templates: [PROGRESS_TEMPLATE]
  sub_workflows: [review_code]
  kb_docs: [architecture/stack.md, architecture/routing.md]
  config_fields: [commands.test, paths.engineering]
---

# 🌱 Meta-Workflow: Implement Plan

## Purpose

The Engineer implements the approved plan: write code, run tests, verify, and document progress.

## Algorithm

```

0a. Pre-flight Gate Check:
   - **Entity-mode resolution:** from the kickoff args, `--task {id}` → `entity_kind="task"`; `--bug {id}` → `entity_kind="bug"`; either sets `record_id={id}`. Every call below uses those two.
   - Run: `forge_preflight({ phase: "implement", {entity_kind}: "{record_id}" })`
   - Exit 1 (gate failed) → print stderr and HALT. Do not proceed; do not attempt to produce the artifact.
   - Exit 2 (misconfiguration) → print stderr and HALT.
   - Exit 0 → continue.

0b. Pipeline Step Guard (user-invoked state check):
   - If `--force` is present in the invocation arguments, skip this step entirely.
   - If `entity_kind == "bug"`, skip this step entirely (bug state is managed by meta-fix-bug.md).
   - Read current task state:
     `forge_store({ command: "read", args: ["task", "{record_id}", "--fields"] }) status`
   - Extract the `status` field from the JSON output.
   - Allowed states for this phase: `plan-approved`.
   - If the current status is NOT in the allowed set:
     Print the following and HALT (do not proceed):
     `× Task {record_id} is in state '{status}' — /forge:review-plan must complete first. To run the full pipeline: /forge:run-task {record_id}`

1. Load Context:
   - Announce your identity via `forge_banner` before any other tool use — `.forge/personas/engineer.md` is already loaded as your system prompt; do not re-read it.
   - store-cli verbs: `read` | `list` | `write` | `emit` | `update-status` | `set-summary` | `describe` | `nlp` | `query` | `delete` — there is no `get`/`set`/`find`. See `_fragments/store-cli-verbs.md` for full notes; run `--help` before improvising.
   - Read the approved PLAN.md
   - Read business domain docs relevant to the task

2. Implementation:
   - Execute plan steps incrementally
   - Perform "compile/check" after each significant change
   - Ensure all new code follows established project patterns

3. Verification:
   - Run syntax verification: {{SYNTAX_CHECK}}
   - Run the test suite: {{TEST_COMMAND}}
     Use exactly that command. Do NOT substitute a bare `python` / `python3` /
     `pytest` — the project interpreter is rarely on `$PATH`, and guessing
     produces `command not found` or `No module named pytest`.
   - Run build if frontend assets modified: {{BUILD_COMMAND}}

4. Documentation:
   - Write PROGRESS.md via `forge_artifact`:
     `forge_artifact({ command:"write", entity:"{entity_kind}", entityId:"{record_id}", artifact:"progress", content:"<markdown>" })`
     Content must include: summary of changes, test evidence (copy of output), files changed manifest.

5. Knowledge Writeback:
   - Update architecture/domain/stack-checklist if discoveries were made
   - Tag updates: `<!-- Discovered during {TASK_ID} — {date} -->`

6. Finalize:
   - Transitions:
     - **Task mode** — legal predecessors are `planned`, `plan-approved`, or `implementing`; target is `implemented`.
       - `planned`        → `implemented` (workflow-prose path — direct)
       - `plan-approved`  → `implementing` → `implemented` (supervisor-review path)
       - Out-of-band escapes (any state): `plan-revision-required`, `code-revision-required`, `blocked`, `escalated`, `abandoned`
       Update status — check current state first:
       - If predecessor is `planned` or `implementing`:
         `forge_store({ command: "update-status", args: ["task", "{record_id}", "status", "implemented"] })`
       - If predecessor is `plan-approved` (two-step mandatory — FSM forbids skipping `implementing`):
         `forge_store({ command: "update-status", args: ["task", "{record_id}", "status", "implementing"] })`
         `forge_store({ command: "update-status", args: ["task", "{record_id}", "status", "implemented"] })`
     - **Bug mode** — NO status write. The bug remains `in-progress` until the commit phase transitions it to `fixed`. Writing `bug.status` here violates `meta-fix-bug.md § Iron Laws #2`.
   - **Do NOT emit a phase event yourself.** The orchestrator owns event emission — it composes the canonical event from runtime telemetry plus the SUMMARY you write next. Subagents that call `store-cli emit` for phase events hallucinate runtime facts (Plan 11 / Slice 2). Write the SUMMARY and return.

7. Emit Summary Sidecar:
   - Write `IMPLEMENTATION-SUMMARY.json` via `forge_artifact`:
     `forge_artifact({ command:"write", entity:"{entity_kind}", entityId:"{record_id}", artifact:"implementation-summary", content:"<JSON>" })`
     JSON shape: `{"objective":"<one sentence>", "key_changes":["<up to 12 bullets>"], "files_changed":["<path>"], "verdict":"n/a", "artifact_ref":"PROGRESS.md"}`
   - `files_changed`: every repo path this phase changed (one `git status --porcelain`); `commit-task.cjs` stages from it.
   - Then link sidecar to store (task mode):
     `forge_store({ command:"set-summary", args:["{record_id}", "implementation"] })`
     Or (bug mode):
     `forge_store({ command:"set-bug-summary", args:["{record_id}", "implementation"] })`
     The sidecar path is auto-resolved from the record's `path` — never pass it.

8. Post-Phase Output Guard: the `outputs` block below is the authoritative enforcer.
   You MUST satisfy it before returning. If PROGRESS.md is missing or too small,
   re-run the relevant step before emitting the complete event.
```

```outputs phase=implement
artifact {engineering}/{sprint}/{task}/PROGRESS.md min=200
require summaries.implementation.verdict == n/a
```

<!-- See _fragments/iron-laws.md for Iron Laws section structure guidance -->
## Iron Laws

- Follow the Algorithm step by step. Execute the approved PLAN.md exactly; do not invent scope or skip steps without updating the plan first.
- Announce your identity via `forge_banner` before any other tool use — `.forge/personas/engineer.md` is already loaded as your system prompt; do not re-read it.
- All store I/O via `forge_store`. Never edit `.forge/store/*.json` directly.
- Run the full test suite before declaring the task implemented. Silent continuation past test failures is never acceptable.

## Store-Write Verification

On a store-write failure (non-zero exit or `PreToolUse` exit 2): fix the JSON and retry, up to 3 times, then halt and escalate. Never set `FORGE_SKIP_WRITE_VALIDATION=1`. Rules: `_fragments/store-write-verification.md`.

## Friction Emit
Emit `type:friction` `{workflow:implement, persona:engineer, issue}` per `_fragments/friction-emit.md`.

<!-- See _fragments/generation-instructions.md for Generation Instructions template -->
## Generation Instructions

- **Workflow Structure:** Strict "Algorithm" block format.
- **Markers (required by `/forge:implement` kickoff shim):** Generated workflow MUST include the "Iron Laws" section, the "Store-Write Verification" section, the literal `forge_store` token, and the `engineer.md` persona path. Missing any → kickoff shim refuses to dispatch.
- **Context Isolation:** Forbid inline execution of complex logic; use the `Agent` tool for sub-tasks.
- **Project Specifics:**
  - Placeholders use the double-brace form ({{SYNTAX_CHECK}}, {{TEST_COMMAND}}, {{BUILD_COMMAND}}) — that is the only form `substitute-placeholders.cjs` matches. A single-brace token is never substituted and reaches the subagent as literal text.
  - Reference project-specific architecture docs by name.
- **Token Reporting:** See `_fragments/finalize.md` — wire via `file_ref:`.
- **Event Emission:** Ensure the "complete" event includes the `eventId` passed by the orchestrator.