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

Spec Writing

by converge-ai-labs in converge-ai-labs/agent-foundation

Author, review, or reorganize accepted technical contracts under spec/, including domain models, lifecycles, APIs, protocols, ownership, security, and compatibility.

Apache-2.0Auto-check passedDevelopment

Install Spec Writing

skills CLI
$ npx skills add converge-ai-labs/agent-foundation --skill spec-writing -a claude-code

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

GitHub CLI
$ gh skill install converge-ai-labs/agent-foundation spec-writing --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/converge-ai-labs/agent-foundation.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/spec-writing .claude/skills/spec-writing && 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
spec-writing
GitHub stars
173
Token cost
~1.1k tokens
SKILL.md length
476 words
Files
3 (incl. references)
Skills in repo
5
Repo updated
First seen
Licence
Apache-2.0

At a glance

Author, review, or reorganize accepted technical contracts under spec/, including domain models, lifecycles, APIs, protocols, ownership, security, and compatibility.

  • Evaluates those contracts
  • SKILL.md covers Establish Authority, Choose the Owner, Write the Contract and Check Consistency and Validate
  • Calls uv, make and git
  • Ordinary implementation details do not require a spec rewrite

What it does

Spec Writing is an agent skill from converge-ai-labs/agent-foundation. Author, review, or reorganize accepted technical contracts under spec/, including domain models, lifecycles, APIs, protocols, ownership, security, and compatibility. Use when a task changes or evaluates those contracts; ordinary implementation details do not require a spec rewrite.

Its SKILL.md is about 1.1k tokens, which your agent loads only when the skill is triggered. The skill folder holds 3 other files, including reference files (for example `references/domain-modeling-and-naming.md` and `references/structure-and-template.md`).

It sits in Development. The repository describes itself as: a13n - An open-source library and self-hosted platform for building and running your own agent systems—with managed agents, memory, sandboxes, computer use, and durable… The licence is Apache-2.0.

When your agent uses it

  • Evaluates those contracts
  • Ordinary implementation details do not require a spec rewrite

Example prompts

  • “/spec-writing”

What it can do on your machine

Read from SKILL.md and the folder at commit b2b8a3e. 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

    Shell commands in SKILL.md call:

    • uv
    • make
    • git

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

  • Network

    No URLs in SKILL.md. Its commands use uv and git, which can reach the network depending on how they are called.

    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

Spec Writing loads about 1.1k tokens when it runs, and up to ~3.1k if it reads all its reference files. Until then it costs about 74 tokens; SKILL.md has 476 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~74
When it runs · the whole SKILL.md, loaded when a task matches
~1.1k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~3.1k

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 converge-ai-labs/agent-foundation at commit b2b8a3e, republished under its Apache-2.0 licence (© converge-ai-labs). 476 words, ~1,092 tokens.

Download SKILL.mdSave it as .claude/skills/spec-writing/SKILL.md (or your agent's skills folder). This skill also uses 2 other files; get the full folder from GitHub.
name
spec-writing
description
Author, review, or reorganize accepted technical contracts under spec/, including domain models, lifecycles, APIs, protocols, ownership, security, and compatibility. Use when a task changes or evaluates those contracts; ordinary implementation details do not require a spec rewrite.

Specification Authoring

Maintain spec/ as the current, internally consistent accepted design. Follow AGENTS.md and CONTRIBUTING.md; Repository Model defines what belongs in specifications.

Establish Authority

Read spec/README.md, the relevant subsystem index/reading path, and the documents that own changed concepts. Search affected terms, states, schemas, and claims across spec/. Reuse context already read.

Use the accepted outcome supplied by the user or concluded discussion. Do not infer a new product or architecture decision from implementation behavior alone. If material alternatives remain unresolved, identify the decision and continue independent authorized work; keep the unsettled design out of spec/. The Issue-to-PR workflow does not itself authorize posting to GitHub.

Choose the Owner

Each durable fact has one owning contract. Update that owner first, then affected summaries, catalogs, reading paths, and incoming links. Preserve established terms and ownership unless the accepted change replaces them.

Read structure-and-template.md when adding or splitting documents, changing ownership/navigation, or choosing a contract's structure. Ordinary edits can retain the existing structure.

Keep platform definition in spec/README.md, subsystem navigation in its README.md, architecture in 00-overview.md, and detailed contracts in cohesive numbered documents. Overviews link to detailed schemas and state machines rather than becoming competing owners. Do not renumber stable files casually or create empty templates/placeholders in spec/.

Write the Contract

Write the resulting design in present tense with testable statements. Explain purpose, ownership, authority, and observable behavior. Include applicable data types, lifecycle/completion boundaries, failure/cancellation/retry/unknown-outcome semantics, security, compatibility, and verifiable invariants.

Read domain-modeling-and-naming.md when adding or reshaping concepts, schemas, identities, lifecycles, or shared terms. Derive models from representative flows before drafting fields or APIs.

Use upstream public primitives when they already own the semantics. Introduce project abstractions only for a stable cross-host or cross-provider contract. Label conceptual Python-like schemas explicitly; distinguish them from serialized wire formats.

Show full SKILL.md (183 more words)Show less

Omit private classes, layouts, hooks, and algorithms that can change without affecting observable behavior, authority, security, or compatibility. Keep accepted trade-offs when they explain the cost of the selected design. Proposals, rejected alternatives, discussion history, progress, tutorials, runbooks, and contributor procedures belong in the surfaces assigned by Repository Model.

Use a diagram or table when it clarifies the contract. Prefer Mermaid for architecture, state, or interaction diagrams, and keep diagrams, schemas, tables, and prose semantically aligned.

Check Consistency and Validate

  • Search changed terminology and update stale summaries, incoming links, and dependent contracts.
  • Verify one owner for each fact, consistent identity/version/state terms, and explicit independent completion boundaries.
  • Distinguish process-local observations from durable facts and transport delivery from execution authority.
  • Make security and compatibility failures explicit. Remove stale text, orphan sections, duplication, and editing residue so the result stands alone without issue history.

For spec-only changes, format the affected files and run:

bash
uv run --locked mdformat --number <changed-spec-files>
make verify
git diff --check -- spec

For accompanying implementation or shared tooling, complete the applicable repository/component gates. Reuse valid results and rerun checks when their inputs change. Report validation and unresolved decisions without claiming that formatting proves semantic correctness.

© converge-ai-labs, Apache-2.0. 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 2 other files (references) in .agents/skills/spec-writing of converge-ai-labs/agent-foundation.

  • SKILL.md
  • references/domain-modeling-and-naming.md
  • references/structure-and-template.md

Open the folder on GitHubat commit b2b8a3e

Compare with similar skills

Spec Writing 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.

Spec Writing compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Spec Writing this skillconverge-ai-labs/agent-foundation173—~1.1kAutomated safety check: PassApache-2.0
Vercel Composition Patternssupabase/supabase111k58 repos~726Automated safety check: PassMIT
Finishing a Development Branchobra/superpowers297k5 repos~1.9kAutomated safety check: PassMIT
Typescript Advanced Typesrolling-scopes/rsschool-app10k25 repos~4.2kAutomated safety check: PassMPL-2.0
PR Babysitteropeninterpreter/openinterpreter69k3 repos~4.2kAutomated safety check: PassApache-2.0
Code Review ChecklistshareAI-lab/learn-claude-code78k4 repos~1.1kAutomated safety check: PassMIT

Similar skills

  • Vercel Composition Patterns

    supabase/supabase

    Official

    React composition patterns that scale. An agent skill from supabase/supabase.

    111k GitHub starsUsed in 58 repos~726 tokens
    DevelopmentAuto-check passed

  • Finishing a Development Branch

    obra/superpowers

    Walks the last step of a branch: confirm tests pass, detect the git environment, ask how to integrate, carry out your choice and clean up the worktree.

    297k GitHub starsUsed in 5 repos~1.9k tokens
    DevelopmentAuto-check passed

  • Typescript Advanced Types

    rolling-scopes/rsschool-app

    Master TypeScript's advanced type system including generics, conditional types, mapped types, template literals, and utility types for building type-safe applications.

    10k GitHub starsUsed in 25 repos~4.2k tokens
    DevelopmentAuto-check passed

  • PR Babysitter

    openinterpreter/openinterpreter

    Watches an open GitHub pull request until it merges, handling review comments, diagnosing CI failures and retrying flaky checks along the way.

    69k GitHub starsUsed in 3 repos~4.2k tokens
    DevelopmentAuto-check passed

  • Code Review Checklist

    shareAI-lab/learn-claude-code

    Reviews code against a five-part checklist covering security, correctness, performance, maintainability and testing, and reports findings in a fixed format.

    78k GitHub starsUsed in 4 repos~1.1k tokens
    DevelopmentAuto-check passed

  • Greploop

    onyx-dot-app/onyx

    Iteratively improves a PR (GitHub), MR (GitLab), or shelved changelist (Perforce) until Greptile gives it a 5/5 confidence score with zero unresolved comments.

    32k GitHub starsUsed in 4 repos~3.3k tokens
    DevelopmentAuto-check passed

More from converge-ai-labs/agent-foundation

  • Audit Repository Consistency

    converge-ai-labs/agent-foundation

    Audit a repository or revision range for spec-code gaps, inconsistent shared concepts, semantic duplication, and removable code.

    173 GitHub stars~1.8k tokensUpdated today
    Auto-check passed

  • Worktree PR

    converge-ai-labs/agent-foundation

    Start a task in a Git worktree under .worktrees/ from freshly fetched origin/main, implement and validate it, then use commit-push-pr to open or update a PR.

    173 GitHub stars~1.2k tokensUpdated today
    Auto-check: notes

  • Commit Push PR

    converge-ai-labs/agent-foundation

    Prepare focused commits, push branches, and create or update GitHub pull requests when requested.

    173 GitHub stars~2.2k tokensUpdated today
    Auto-check passed

  • Uiux Polish

    converge-ai-labs/agent-foundation

    Build and refine frontend pages and interactions with strong visual taste, proactive polish, and lightweight iteration.

    173 GitHub stars~580 tokensUpdated today
    Auto-check passed

Categories

Questions about Spec Writing

What does Spec Writing do?

Author, review, or reorganize accepted technical contracts under spec/, including domain models, lifecycles, APIs, protocols, ownership, security, and compatibility. Spec Writing is an agent skill from converge-ai-labs/agent-foundation. Author, review, or reorganize accepted technical contracts under spec/, including domain models, lifecycles, APIs, protocols, ownership, security, and compatibility.

When should I use Spec Writing?

Spec Writing fits situations like: evaluates those contracts; ordinary implementation details do not require a spec rewrite.

How do I install Spec Writing in Claude Code?

Run `npx skills add converge-ai-labs/agent-foundation --skill spec-writing -a claude-code`. Or copy the skill folder (.agents/skills/spec-writing in converge-ai-labs/agent-foundation) into .claude/skills/spec-writing in your project. Claude Code loads it when a task matches its description.

How do I install Spec Writing in Codex?

Run `npx skills add converge-ai-labs/agent-foundation --skill spec-writing -a codex`. Or copy the skill folder (.agents/skills/spec-writing in converge-ai-labs/agent-foundation) into .agents/skills/spec-writing in your project. Codex loads it when a task matches its description.

Can I use Spec Writing 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 converge-ai-labs/agent-foundation --skill spec-writing -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/spec-writing, .gemini/skills/spec-writing, .github/skills/spec-writing and .opencode/skills/spec-writing in your project.

What does Spec Writing need to run?

Going by SKILL.md and its folder, Spec Writing needs the command-line tools its instructions call (uv, make and git).

Does Spec Writing access the network?

SKILL.md contains no URLs. Its commands use uv and git, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Spec Writing 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 Spec Writing use?

Spec Writing is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Spec Writing use?

About 1.1k tokens (SKILL.md is roughly 4.4k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 2k tokens, read only when the agent opens those files.

What are the alternatives to Spec Writing?

Skills that share tags, products or a category with Spec Writing: Vercel Composition Patterns (supabase/supabase, 111k stars), Finishing a Development Branch (obra/superpowers, 297k stars), Typescript Advanced Types (rolling-scopes/rsschool-app, 10k stars) and PR Babysitter (openinterpreter/openinterpreter, 69k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Spec Writing?

converge-ai-labs (a GitHub organization) maintains it in converge-ai-labs/agent-foundation, which has 173 GitHub stars. The repository holds 5 skills in this directory. The repository was last updated on October 11, 2026.

Source: converge-ai-labs/agent-foundation on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.