Skip to content
Skills Library
Programming · Building & coding

dev-implement

/dev-implement

Execute an approved plan task by task with a check after each file, log every deviation against the acceptance criteria, and apply stop rules for build and type errors. Use after the plan is approved, or when the user says "làm đi" on a planned change, or when a build/typecheck/lint keeps failing.

Repository
bdt-kit ↗
Licence
MIT
Last repository change
30.10.2026

Assessment and security

90/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content25 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 30.10.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Execute an approved plan task by task with a check after each file, log every deviation against the acceptance criteria, and apply stop rules for build and type errors. Use after the plan is approved, or when the user says "làm đi" on a planned change, or when a build/typecheck/lint keeps failing.

Triggers from the description

làm đi
Original author description

Execute an approved plan task by task with a check after each file, log every deviation against the acceptance criteria, and apply stop rules for build and type errors. Use after the plan is approved, or when the user says "làm đi" on a planned change, or when a build/typecheck/lint keeps failing.

How it works

  1. Loop per task
  2. Deviations
  3. Stop rules (build, type and lint errors)
  4. Hygiene before leaving this step
  5. Then
Original SKILL.md headings · 31 lines, 269 words
SKILL.md preview
---
name: dev-implement
description: Execute an approved plan task by task with a check after each file, log every deviation against the acceptance criteria, and apply stop rules for build and type errors. Use after the plan is approved, or when the user says "làm đi" on a planned change, or when a build/typecheck/lint keeps failing.
---

# dev-implement

**Input:** an approved plan.
**Output:** the code change, deviations logged in `report.md`, checks run after each task.

Plan: `.bdt/work/<TICKET-ID>/plan.md`; deviations go into `report.md` beside it.

## Loop per task
1. Re-read the mirror file:line before writing.
2. Make the smallest change for this task. Follow the CONV rules the task lists and the stack pack's idioms (`.bdt/dev/`). Match surrounding naming, comments and idiom.
3. Run the task's Validate step (at least typecheck or lint of the touched files), and the Check of each listed CONV rule that is a grep or a lint rule.
4. Tick the task. If you did something the plan did not say, log it: what changed, why, which AC.

## Deviations
- A deviation that changes behaviour or scope is shown to the user before continuing.
- An AC that turns out impossible: mark `[revised]` in the brief, add a question, tell the user.

## Stop rules (build, type and lint errors)
Fix one error at a time and re-run. Stop and report when:
- the same error survives 3 attempts;
- a fix creates more errors than it removes;
- the cause is architectural (wrong layer, missing contract) rather than local.
Never silence a check to get green: no type escapes, ignore comments, disabled rules or skipped git hooks. The stack pack lists the forms for this stack.

## Hygiene before leaving this step
No debug output (the stack pack lists the patterns), no commented-out code, no unused imports, no stray files, formatter run on touched files only, lockfile unchanged unless dependencies changed.

## Then
Run `qa-verify`. Do not commit yet.

Versions

The database has no other skills named dev-implement.

Database snapshot: 13.11.2026 · package version fada5a08abde (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100execute
    Execute an approved implementation plan from .ai/plans/ task by task, running each task's verification and committing per task via the check-and-commit gate. Use when asked to "execute the plan", "implement the plan", or after /plan approval. Do not use without an approved plan…
  • 90build-from-plan
    Implements an approved plan from the feature worktree with scoped tests and verification. Use when the user says implement the plan, execute tasks, or build the feature.
  • 90dev-build
    Build an approved plan task by task with a fresh implementer and a fresh reviewer per task, a ledger, capped fix rounds, converge, a final review and ship. Use for "build this plan", "run the plan", "continue the build", or when dev reaches the build step.
  • 90evanflow-executing-plans
    Execute a written implementation plan task-by-task with inline verification. Stop and ask on blockers. No forced sub-skill chains. Use when a plan exists and you're ready to implement.
  • 100implement-plan
    Execute an approved implementation plan from the configured plans directory. Use when the user wants to implement a plan phase by phase with verification.
  • 90executing
    Runs a written plan task by task with checks and a final review. Use when a plan exists and the user says go, implement it, execute the plan, or build the planned feature. Not for writing plans, single edits or reviews.

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.