New: 98,000+ skills indexed from GitHub repos with 100+ stars · every skill safety-checkedBrowse top skillsTop skills
Esc to closeBrowse all skills
Submit a skillBrowse skills
Submit a skillBrowse skills

Agent skill

Project Skill Review

by TwiTech-LAB in TwiTech-LAB/devchain

Review the skills stored for a project. An agent skill from TwiTech-LAB/devchain.

MITAuto-check passedData & Analytics

Install Project Skill Review

skills CLI
$ npx skills add TwiTech-LAB/devchain --skill project-skill-review -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install TwiTech-LAB/devchain project-skill-review --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/TwiTech-LAB/devchain.git skills-src && mkdir -p .claude/skills && cp -r skills-src/apps/local-app/skills/project-skill-review .claude/skills/project-skill-review && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
project-skill-review
GitHub stars
101
Token cost
~2k tokens
SKILL.md length
1,153 words
Files
2
Skills in repo
6
Repo updated
First seen
Licence
MIT

At a glance

Review the skills stored for a project. An agent skill from TwiTech-LAB/devchain.

  • Works in 12 steps: Offer a sync → Read the statistics first → Read the project specifications → …
  • The skill list is too long
  • SKILL.md covers Step 0 — Offer a sync, Step 1 — Read the statistics…, Step 2 — Read the project… and Step 3 — Judge the statistics, plus 11 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Project Skill Review is an agent skill from TwiTech-LAB/devchain. Review the skills stored for a project. Read the project specifications and the skill usage statistics. Classify each enabled skill as Keep, Disable, Protected, or Unsure. Classify each disabled skill as Enable or Stay disabled. Propose source bundles, source disables, and a skill sync. Apply a change only after the user approves it. Use when the skill list is too long, when agents pick wrong skills, or when the user asks to enable, disable, prune, or sync skills. Triggers: review project skills, enable disabled…

Its SKILL.md is about 2k tokens, which your agent loads only when the skill is triggered. The skill folder holds 1 other file (for example `README.md`).

It sits in Data & Analytics, covering Statistics. The repository describes itself as: Run a team of AI coding agents. On your machine. Local-first orchestration for Claude Code, Codex, OpenCode, Antigravity, and Copilot. The licence is MIT.

When your agent uses it

  • The skill list is too long
  • Agents pick wrong skills
  • The user asks to enable

Example prompts

  • “/project-skill-review”

Workflow steps

12 steps, taken from the step headings in SKILL.md.

  1. Offer a sync
  2. Read the statistics first
  3. Read the project specifications
  4. Judge the statistics
  5. List all stored skills
  6. Find the protected skills
  7. Classify each enabled skill
  8. Classify each disabled skill
  9. Build source bundles
  10. Propose source disables
  11. Handle globally disabled sources
  12. Report and ask

What it can do on your machine

Read from SKILL.md and the folder at commit 172b47c. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Project Skill Review loads about 2k tokens when it runs. Until then it costs about 161 tokens; SKILL.md has 1,153 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~161
When it runs · the whole SKILL.md, loaded when a task matches
~2k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from TwiTech-LAB/devchain at commit 172b47c, republished under its MIT licence (© TwiTech-LAB). 1,153 words, ~2,010 tokens.

Download SKILL.mdSave it as .claude/skills/project-skill-review/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.
name
project-skill-review
description
Review the skills stored for a project. Read the project specifications and the skill usage statistics. Classify each enabled skill as Keep, Disable, Protected, or Unsure. Classify each disabled skill as Enable or Stay disabled. Propose source bundles, source disables, and a skill sync. Apply a change only after the user approves it. Use when the skill list is too long, when agents pick wrong skills, or when the user asks to enable, disable, prune, or sync skills. Triggers: review project skills, enable disabled skills, disable unused skills, shorten the skill list, sync skills, skill cleanup, project skill review.
displayName
Project Skill Review
version
0.2.1
license
MIT

Project Skill Review

Review all skills stored for this project. Propose enables, disables, source bundles, and source disables. The goal is a short, relevant skill list, so agents pick the right skill faster.

Never write without explicit user approval. Never disable a protected skill. Never write to a globally disabled source.

Step 0 — Offer a sync

  1. Propose devchain_skills_sync for the sources that matter to this review. Prefer a named source. Use a full sync only when no name fits.
  2. Run the sync only after the user approves it. Continue when the user declines.
  3. Treat the outcome as "catalog not confirmed fresh" when a timeout happens, when status is already_running, when failed is above 0, or when errors is not empty. Say so in the report.
  4. Call the skill list "the last stored catalog".

Step 1 — Read the statistics first

Call devchain_skills_usage_stats before any devchain_get_skill call. Each devchain_get_skill call creates a usage event. A call during the review changes the data that this review reads.

Step 2 — Read the project specifications

  1. Read the docs entry point. Use docs/AGENTS.md. Use docs/README.md when docs/AGENTS.md is absent.
  2. Read the root README.
  3. Read the package manifests.

Find the main languages and the main frameworks. Write a list of the areas of usage. Example: "backend: NestJS and SQLite", "UI: React and Vite".

Step 3 — Judge the statistics

  1. The span is lastEventAt − firstEventAt. Do not use the window that the request sent.
  2. The statistics are sufficient only when both conditions hold: the span is 30 days or more, and totalEvents is 100 or more.
  3. When the statistics are not sufficient, say so in the report. Decide from the specifications only.
  4. State in the report: usage means successful devchain_get_skill loads. It does not prove that a skill helped.

Step 4 — List all stored skills

Call devchain_list_skills with includeDisabled: true. The response is the last stored catalog. Each item carries four flags:

  • disabled: the effective state. It is true when any of the other three blocks applies.
  • skillDisabled: the project disabled this skill.
  • sourceProjectEnabled: the project enables the source. The default is true.
  • sourceGloballyEnabled: the settings enable the source. The default is true.

A skill is enabled when disabled is false. A skill is disabled when disabled is true.

Step 5 — Find the protected skills

From epicReferences
  1. Choose the status labels that mean closed. Examples: Done, Archive.
  2. Name these labels in the report.
  3. A slug is protected when an epic that is not closed references it.
  4. Treat an ambiguous label as open.
From the prompts
  1. Call devchain_list_prompts.
  2. Call devchain_get_prompt for each prompt.
  3. A prompt protects a skill when it contains the slug of that skill or the bare name of that skill.
On failure

When a list call or a fetch call fails, say in the report that the protection is incomplete. Keep every skill that you could not check.

Step 6 — Classify each enabled skill

The usage identity is the exact full slug. Review each skill with disabled: false. Do not skip a skill. Put each skill in exactly one class:

  1. Protected: an open epic or a prompt requires the skill. Keep it.
  2. Keep: the skill has usage in the statistics. Or the skill matches an area of usage.
  3. Disable: all three conditions hold: the skill has no usage and the usage list is complete, so no usage means zero use; the skill matches no area of usage; the skill is not protected.
  4. Unsure (keep): all other cases. Example: a skill has no usage, but a skill with the same name in another source has usage.

Skip skills with sourceGloballyEnabled: false here. Step 10 covers them.

Step 7 — Classify each disabled skill

Review each skill with disabled: true and sourceGloballyEnabled: true. This covers skills the project disabled and skills of sources the project disabled. Apply the classes in this order:

  1. Enable (required): an open epic or a prompt references the skill.
  2. Enable: the skill clearly matches an area of usage, and no enabled skill has the same slug name segment.
  3. Unsure (stay disabled): the skill has past usage without an area match. Or an enabled skill with the same name segment exists.
  4. Stay disabled: all other cases.

Never propose Enable because of past usage alone.

Show full SKILL.md (430 more words)Show less

Step 8 — Build source bundles

A source bundle turns one project-disabled source on, without its noise. Build one bundle for each source with sourceProjectEnabled: false and sourceGloballyEnabled: true:

  1. Propose "enable source X for this project" only when X holds at least one Enable-class skill.
  2. Add a companion disable for every other skill of X that is not Enable-class and has skillDisabled: false.
  3. The user approves or declines the whole bundle. When the user declines, X stays disabled.
  4. Apply in this order: first the companion disables (devchain_skills_set_enabled with enabled: false), then the source enable (devchain_skills_set_source_enabled with enabled: true).
  5. When the disable call returns a nonempty notFound or fails, do not enable the source. Report the bundle as not applied.

Step 9 — Propose source disables

  1. Compute these proposals from the final approved state. Count the approved bundles and skill enables first.
  2. Propose "disable source X for this project" when every enabled skill of X is Disable-class.

Step 10 — Handle globally disabled sources

  1. A skill with sourceGloballyEnabled: false is read-only for MCP. Its stored content can be stale, because disabled sources do not sync.
  2. Never write to these skills and never write to their source.
  3. The report may recommend: "enable source X in Settings, then sync X".
  4. When the user enables such a source in Settings, propose devchain_skills_sync for that source.

Step 11 — Report and ask

Show the report in eight groups: Enable, Disable, Source bundles, Source disables, Unsure, Keep, Protected, Globally disabled. The Globally disabled group holds recommendations only. Give one short reason per skill. Show the statistics window, the sufficiency result, the closed labels, and the sync result.

The user approves each group separately. The user can remove items from a group.

Step 12 — Apply only after approval

  1. Apply in this order: source bundles, skill enables, skill disables, source disables.
  2. Report updatedCount and the counts of unchanged and notFound for each write. A nonempty notFound means an incomplete application. Say so.
  3. The changed slugs of a write are the requested slugs minus unchanged and notFound. Build each undo call from the changed slugs only. Use only the sources that changed. A bundle undo disables the source again, then re-enables its changed companion skills.
  4. Give the undo calls in the report.

Local skill edits

After an edit of a local skill, edit the authoritative local source folder, not the materialized contentPath copy. Then propose devchain_skills_sync for that source.

Rules

  • Never write without explicit user approval.
  • Never disable a protected skill.
  • Never write to a globally disabled source.

© TwiTech-LAB, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 1 other file in apps/local-app/skills/project-skill-review of TwiTech-LAB/devchain.

  • SKILL.md
  • README.md

Open the folder on GitHubat commit 172b47c

Compare with similar skills

Project Skill Review next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Project Skill Review compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Project Skill Review this skillTwiTech-LAB/devchain101—~2kAutomated safety check: PassMIT
Sandbox Benchvercel/next.js143k—~4.1kAutomated safety check: PassMIT
Statistical Analysisspacering-net/codeg3.9k3 repos~5kAutomated safety check: PassMIT
StatsmodelszLanqing/codex-claude-academic-skills4.7k15 repos~4.9kAutomated safety check: PassBSD-3-Clause
AI Daily DigestvigorX777/ai-daily-digest1.6k—~1.3kAutomated safety check: PassNone
Statistical Powerspacering-net/codeg3.9k1 repos~3.6kAutomated safety check: NotesMIT

Similar skills

  • Sandbox Bench

    vercel/next.js

    Official

    Benchmark React or Next.js changes on Vercel Sandbox VMs with paired A/B statistics: react PR/commit vs base, or Next.js PR/commit vs base, measured end-to-end through the bench/render-pipeline app…

    143k GitHub stars~4.1k tokensUpdated today
    Data & AnalyticsAuto-check passed

  • Statistical Analysis

    spacering-net/codeg

    Guided statistical analysis for research data - test selection, assumption checking, effect sizes, power analysis, Bayesian alternatives, and APA-formatted reporting.

    3.9k GitHub starsUsed in 3 repos~5k tokens
    Data & AnalyticsAuto-check passed

  • Statsmodels

    zLanqing/codex-claude-academic-skills

    Statistical models library for Python. An agent skill from zLanqing/codex-claude-academic-skills.

    4.7k GitHub starsUsed in 15 repos~4.9k tokens
    Data & AnalyticsAuto-check passed

  • AI Daily Digest

    vigorX777/ai-daily-digest

    Fetches RSS feeds from 90 top Hacker News blogs (curated by Karpathy), uses AI to score and filter articles, and generates a daily digest in Markdown with Chinese-translated titles, category…

    1.6k GitHub stars~1.3k tokensUpdated 7 mo ago
    Data & AnalyticsAuto-check passed

  • Statistical Power

    spacering-net/codeg

    Sample-size and statistical power calculations for planning studies.

    3.9k GitHub starsUsed in 1 repo~3.6k tokens
    Data & AnalyticsAuto-check: notes

  • Agent Session Monitor

    higress-group/higress

    Real-time agent conversation monitoring - monitors Higress access logs, aggregates conversations by session, tracks token usage.

    9.5k GitHub stars~3.3k tokensUpdated yesterday
    Data & AnalyticsAuto-check passed

More from TwiTech-LAB/devchain

  • Asd Ste100 Skill

    TwiTech-LAB/devchain

    Write or rewrite English with ASD-STE100 Simplified Technical English rules: one meaning per word, active voice, simple tenses, one instruction per sentence.

    101 GitHub stars~1.2k tokensUpdated 2 days ago
    Auto-check passed

  • Vm Provisioning

    TwiTech-LAB/devchain

    Prepare a DevChain remote VM for development of a project. An agent skill from TwiTech-LAB/devchain.

    101 GitHub stars~1.8k tokensUpdated 2 days ago
    Auto-check: notes

  • Code Simplifier

    TwiTech-LAB/devchain

    Review a change set for reuse, simplification, efficiency, and altitude problems, then apply the fixes while preserving exact behavior.

    101 GitHub stars~2k tokensUpdated 2 days ago
    Auto-check passed

  • Create Commit

    TwiTech-LAB/devchain

    Create a git commit whose message records why the change exists, not just what changed.

    101 GitHub stars~1.2k tokensUpdated 2 days ago
    Auto-check passed

  • Improve Codebase Architecture

    TwiTech-LAB/devchain

    Plan architecture improvements and refactoring for a codebase: scan for deepening opportunities (shallow modules, leaky seams, low-leverage interfaces), present candidates as a markdown report with…

    101 GitHub stars~3k tokensUpdated 2 days ago
    Auto-check passed

Categories

Questions about Project Skill Review

What does Project Skill Review do?

Review the skills stored for a project. An agent skill from TwiTech-LAB/devchain. Project Skill Review is an agent skill from TwiTech-LAB/devchain. Review the skills stored for a project.

When should I use Project Skill Review?

Project Skill Review fits situations like: the skill list is too long; agents pick wrong skills; the user asks to enable.

How do I install Project Skill Review in Claude Code?

Run `npx skills add TwiTech-LAB/devchain --skill project-skill-review -a claude-code`. Or copy the skill folder (apps/local-app/skills/project-skill-review in TwiTech-LAB/devchain) into .claude/skills/project-skill-review in your project. Claude Code loads it when a task matches its description.

How do I install Project Skill Review in Codex?

Run `npx skills add TwiTech-LAB/devchain --skill project-skill-review -a codex`. Or copy the skill folder (apps/local-app/skills/project-skill-review in TwiTech-LAB/devchain) into .agents/skills/project-skill-review in your project. Codex loads it when a task matches its description.

Can I use Project Skill Review in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add TwiTech-LAB/devchain --skill project-skill-review -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/project-skill-review, .gemini/skills/project-skill-review, .github/skills/project-skill-review and .opencode/skills/project-skill-review in your project.

What does Project Skill Review need to run?

SKILL.md names no scripts, command-line tools or credentials: Project Skill Review is instructions for the agent only.

Does Project Skill Review access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Project Skill Review safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Project Skill Review use?

Project Skill Review is published under the MIT licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Project Skill Review use?

About 2k tokens (SKILL.md is roughly 8k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Project Skill Review?

Skills that share tags, products or a category with Project Skill Review: Sandbox Bench (vercel/next.js, 143k stars), Statistical Analysis (spacering-net/codeg, 3.9k stars), Statsmodels (zLanqing/codex-claude-academic-skills, 4.7k stars) and AI Daily Digest (vigorX777/ai-daily-digest, 1.6k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Project Skill Review?

TwiTech-LAB (a GitHub organization) maintains it in TwiTech-LAB/devchain, which has 101 GitHub stars. The repository holds 6 skills in this directory. The repository was last updated on October 7, 2026.

Source: TwiTech-LAB/devchain on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.