/cardume-core:implement — Build task by task
$ARGUMENTS
Read tasks.md and its spec.md. If no path was given, use the most recently modified
docs/specs/*/tasks.md.
Before starting
Check both status lines. Start only when tasks.md says Status: approved and its spec says
approved (or implemented, for follow-up tasks). Otherwise stop and tell the user exactly what is
missing — e.g. "spec.md is draft with 2 [NEEDS CLARIFICATION] markers; tasks.md has no approval" —
and the command that fixes it. Building on unapproved plans is the most common source of rework.
Then loop over unchecked tasks in order, respecting Depends on.
Per task
- Delegate to a fresh subagent (the task's
Agent). Give it only what it needs: the task block,
the requirement rows it covers from spec.md, the relevant Design notes and the files list. Fresh
context per task keeps earlier mistakes from leaking forward.
Tasks with no dependency between them may run in parallel; give each its own worktree
(isolation: worktree) when they touch overlapping files.
- Test first for logic. For tasks that change logic, the subagent adds or names the test for the
requirement, runs it before the fix and shows it failing, then shows it passing after. A test that
never failed may not test the change.
- Verify. Run the task's
Verify command yourself and read the output. A task is done only
when the command shows the expected result — not when the subagent says so.
- Independent review. Send the diff, the task block and its requirement rows to a separate
reviewer subagent that did not write the code:
test-engineer for test-only tasks, otherwise a
general-purpose subagent using the code-review-checklist skill. Ask for two verdicts:
- Spec compliance: does it do what the requirement rows say, nothing more, nothing missing?
- Code quality: correctness, tests,
docs/standards.md conventions.
Each verdict is pass or fail with specific findings (file:line, problem, fix).
- Fix loop. On a fail, send the findings back to the implementing subagent and review again.
Cap it at 2 rounds; if the task still fails, stop and show the user the findings and what was tried.
- Record. Tick the checkbox and append under the task:
- Evidence: <command> → <result summary> and, for logic tasks,
- Test-first: <test> failed before (<message>) → passes after.
When something goes wrong
- Verification fails twice for the same reason → stop the loop, show the user the failure and what
was tried; the task or the spec probably needs adjusting.
- The task reveals the spec is wrong or incomplete → stop and propose the spec change; do not
silently implement something different from the approved requirement.
Finish
When all tasks are ticked, run cardume-checklist . and a final whole-branch review with the
review skill (/cardume-core:review), since per-task reviews miss problems between tasks. Report
both results, then suggest /cardume-core:verify docs/specs/<slug>/spec.md.
1---2name: implement3description: "Builds an approved tasks.md one task at a time, with a fresh subagent, a verification command and a separate reviewer. Use to implement level-2/3 work once its tasks are approved."4argument-hint: "[path to tasks.md] [task id to start from]"5compatibility: "Claude Code with the Cardume plugins (uses cardume-* commands, subagents or plugin paths); the guidance itself is portable."6metadata:7 source: https://skillmd.com/skills/jeffersonvalente/implement8---910# /cardume-core:implement — Build task by task1112$ARGUMENTS1314Read `tasks.md` and its `spec.md`. If no path was given, use the most recently modified15`docs/specs/*/tasks.md`.1617## Before starting1819Check both status lines. Start only when tasks.md says `Status: approved` and its spec says20`approved` (or `implemented`, for follow-up tasks). Otherwise stop and tell the user exactly what is21missing — e.g. "spec.md is draft with 2 [NEEDS CLARIFICATION] markers; tasks.md has no approval" —22and the command that fixes it. Building on unapproved plans is the most common source of rework.2324Then loop over unchecked tasks in order, respecting `Depends on`.2526## Per task27281. **Delegate to a fresh subagent** (the task's `Agent`). Give it only what it needs: the task block,29 the requirement rows it covers from spec.md, the relevant Design notes and the files list. Fresh30 context per task keeps earlier mistakes from leaking forward.31 Tasks with no dependency between them may run in parallel; give each its own worktree32 (`isolation: worktree`) when they touch overlapping files.332. **Test first for logic.** For tasks that change logic, the subagent adds or names the test for the34 requirement, runs it before the fix and shows it failing, then shows it passing after. A test that35 never failed may not test the change.363. **Verify.** Run the task's `Verify` command yourself and read the output. A task is done only37 when the command shows the expected result — not when the subagent says so.384. **Independent review.** Send the diff, the task block and its requirement rows to a separate39 reviewer subagent that did not write the code: `test-engineer` for test-only tasks, otherwise a40 general-purpose subagent using the `code-review-checklist` skill. Ask for two verdicts:41 - **Spec compliance:** does it do what the requirement rows say, nothing more, nothing missing?42 - **Code quality:** correctness, tests, `docs/standards.md` conventions.43 Each verdict is pass or fail with specific findings (`file:line`, problem, fix).445. **Fix loop.** On a fail, send the findings back to the implementing subagent and review again.45 Cap it at 2 rounds; if the task still fails, stop and show the user the findings and what was tried.466. **Record.** Tick the checkbox and append under the task:47 ` - Evidence: <command> → <result summary>` and, for logic tasks,48 ` - Test-first: <test> failed before (<message>) → passes after`.4950## When something goes wrong5152- Verification fails twice for the same reason → stop the loop, show the user the failure and what53 was tried; the task or the spec probably needs adjusting.54- The task reveals the spec is wrong or incomplete → stop and propose the spec change; do not55 silently implement something different from the approved requirement.5657## Finish5859When all tasks are ticked, run `cardume-checklist .` and a final whole-branch review with the60`review` skill (`/cardume-core:review`), since per-task reviews miss problems between tasks. Report61both results, then suggest `/cardume-core:verify docs/specs/<slug>/spec.md`.