adev:implement

A tool for carrying out a written software implementation plan through separate specialist agents. It requires test-first development, where tests are written before the code, and reviews each task in two stages.

In plain words
What is it for?
Use it after planning a feature to implement individual tasks, route work to relevant specialists, run tasks in parallel when files do not overlap, or re-run a selected task.
Why use it?
It helps turn a plan into checked, incremental changes while catching both missed requirements and code-quality problems after each task.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/agentic-development/adev-plugin/implement
Any agent
npx skills add agentic-development/adev-plugin --skill implement
Clone the repo
git clone --depth 1 https://github.com/agentic-development/adev-plugin

Made for: Claude Code, Codex.

Per session 42 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 15,157 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00042 $0.15157
Opus 5 $0.00021 $0.07579
Sonnet 5 $0.00008 $0.03031
Haiku 4.5 $0.00004 $0.01516

Measured 2d ago against content hash dd8c51d3ec3a, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

adev:implement scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

providers/codex/skills/implement/SKILL.md · 875 lines

How it starts

The opening of the file, as written. The whole thing — 875 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Implement Plan

Execute an implementation plan by dispatching a fresh subagent per task, routing to domain specialists when applicable, enforcing TDD, and running 2-stage review (spec compliance then code quality) after each task.

Arguments

  • <plan-path>: path to the plan file (required). Usually .context-index/specs/features/<module>/<spec-slug>.plan.md.
  • --task <N>: execute only task N (useful for re-running a single task after a fix)
  • --dry-run: show routing decisions and specialist matches without executing
  • --no-infra: skip infrastructure preflight checks (user-only — the agent must never set this flag)
  • --verbose: disable silent execution for per-task subagents. Includes VERBOSE: true in subagent prompts so they narrate each step. Useful for debugging task failures.
  • --parallel: run file-disjoint task groups concurrently in adev-managed worktrees instead of strictly serially (see Step 2.5). Falls back to serial when the plan has no usable ## Parallelization section.
  • --fresh: with --parallel, on a re-run collision auto-remove the retained worktree (adev worktree remove --force) and proceed, instead of aborting the group with RERUN_COLLISION. No effect without --parallel.
  • --no-batch: force solo dispatch for every task, restoring today's strict one-subagent-per-task behavior. Rejected with CONFLICTING_BATCH_FLAGS when combined with --parallel (--parallel's unit of dispatch is already the group — the two flags would disagree about what "batching off" means).
  • --max-batch <n>: per-run override of implement.max_batch_size (default 4). 1 is equivalent to --no-batch.
  • --review-cycles <n>: per-run override of implement.max_review_cycles (default 3), forwarded to every adev implement resolve-depth call (the verb is the only validator).

Prerequisites

Before starting, verify all five conditions. If any fails, stop and tell the user what to fix.

  1. No conflicting batch flags. If both --no-batch and --parallel were passed to this skill invocation, stop immediately with CONFLICTING_BATCH_FLAGS — do not proceed to Step 1, Step 2, or Step 2.5. This check runs here, unconditionally, precisely because it must fire regardless of which of those steps the run would otherwise take: --parallel present routes to Step 2.5's group dispatch instead of Step 2's per-task loop, and Step 2's own "Batch resolution" paragraph (where adev implement batches is invoked) never executes on that path — so the flag conflict would otherwise go undetected on exactly the run where the operator most needs to hear about it. Report the same message adev implement batches itself would give: "CONFLICTING_BATCH_FLAGS: --no-batch and --parallel are mutually exclusive — --parallel's unit of dispatch is already the group. Drop one flag."
  2. Plan exists. The plan file must exist and be readable.
  3. Context Index exists. .context-index/ must be present with constitution.md and manifest.yaml.
  4. Plan step gate. As the FIRST action in the skill — before reading the plan file or loading context — gate on the prior step via the lifecycle log, then emit the step-started event:

Read the full file on GitHub · 875 lines

Files

What ships with it

10 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 875 lines · 42 tokens per session scan A dd8c51d3ec3a

Subscribe to this mod's changes

adev:implement is a skill published in the GitHub repository agentic-development/adev-plugin (11 stars, last pushed 2d ago), licensed MIT. It adds 42 tokens to every session and 15,157 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

next-cache-components-adoption

Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…

vercel/next.js · 95 tokens

babysit-pr

Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…

openai/codex · 114 tokens

imagegen

Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…

openai/codex · 113 tokens

cpu-profile-analysis

Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…

microsoft/vscode · 71 tokens

next-cache-components-optimizer

Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…

vercel/next.js · 170 tokens