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.
npx agentmods add skills/context4ai/workflow/implement-by-plannpx skills add context4ai/workflow --skill implement-by-plangit clone --depth 1 https://github.com/context4ai/workflowWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00034 | $0.02718 |
| Opus 5 | $0.00017 | $0.01359 |
| Sonnet 5 | $0.00007 | $0.00544 |
| Haiku 4.5 | $0.00003 | $0.00272 |
Grade A, and why
implement-by-plan 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 yesterday.
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.
How it starts
The opening of the file, as written. The whole thing — 269 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/implement-by-plan - Execute Assigned Agent Tasks from the Plan
Execute the tasks assigned to a specific Agent number from the Plan (Code-layer graph).
Design Workspace (design-dir)
The default workspace for design documents is design/. On startup the Agent checks:
design/directory exists --> usedesign/as<design-dir>design/does not exist --> pause and ask the user for the design workspace path, explaining the accepted directory structure:<design-dir>/<version>/(prd|plan).md Examples: design/v1.0.1/prd.md, design/v0.5.0/plan.md
design/archive/: Archived or inactive design documents. When exploring existing graph structures, treat as low-priority reference -- read active version directories first and fall back to archive only when relevant context cannot be found in active versions.
All paths below use <design-dir> to refer to this directory.
TL;DR -- Hard Rules
- PRD is the sole authority: Feature definitions defer to the PRD; the Plan is the execution path
- Only work on the Phase assigned to this Agent -- no boundary crossing. Role is determined by the Plan assignment table
Rolecolumn (task/e2e); do not guess - Acceptance evidence goes in
agent-<n>-status.md-- do not pollute the Plan file - Plan writeback: Update the Plan when deviating, but spec content changes must never be edited in place -- must use
[-Pn]+[+Pnew]+supersedes - Breaking downstream packages: Whoever breaks it fixes it; append a repair Phase at the end of the Plan
- verify exemption: For cross-package breaking changes, typecheck may be deferred (
[verify:deferred-typecheck]), but unit tests for the current package must pass - Do not touch other Agents' code/tests; if issues are found, record them only -- do not fix
Agent Responsibility Boundaries
Must Do
- Complete all guide steps for Nodes marked [new]/[modified]/[delete] in the Phase assigned to this Agent in the Plan
- Write tests for new code (corresponding to acceptance criteria marked
[test]in[Pn][spec]) - Pass quality gates (per the verify level noted in the Plan)
- Update the Plan file when implementation deviates from the Plan
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.
- yesterday First seen · 269 lines · 34 tokens per session scan A fb3835ad303a
implement-by-plan is a skill published in the GitHub repository context4ai/workflow (4 stars, last pushed 4mo ago), licensed MIT. It adds 34 tokens to every session and 2,718 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 2026-08-31.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
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…
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…
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…
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…
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…