Intent Layer
crafter-station/skills
Set up hierarchical Intent Layer (AGENTS.md files) for codebases.
Scans a repository for real evidence and proposes, or on request writes, a small stack of root and scoped AGENTS.md files with validation commands and generated-file boundaries.
Install AGENTS.md Scaffold
$ npx skills add majiayu000/spellbook --skill agentsmd-scaffold -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install majiayu000/spellbook agentsmd-scaffold --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/majiayu000/spellbook.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/agentsmd-scaffold .claude/skills/agentsmd-scaffold && rm -rf skills-srcUse ~/.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/
Install the "agentsmd-scaffold" agent skill from https://github.com/majiayu000/spellbook/tree/main/skills/agentsmd-scaffold into .claude/skills/agentsmd-scaffold/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "agentsmd-scaffold", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/majiayu000/spellbook/tree/main/skills/agentsmd-scaffoldType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add majiayu000/spellbook --skill agentsmd-scaffold -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install majiayu000/spellbook agentsmd-scaffold --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/majiayu000/spellbook.git skills-src && mkdir -p .agents/skills && cp -r skills-src/skills/agentsmd-scaffold .agents/skills/agentsmd-scaffold && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "agentsmd-scaffold" agent skill from https://github.com/majiayu000/spellbook/tree/main/skills/agentsmd-scaffold into .agents/skills/agentsmd-scaffold/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "agentsmd-scaffold", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add majiayu000/spellbook --skill agentsmd-scaffold -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install majiayu000/spellbook agentsmd-scaffold --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/majiayu000/spellbook.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/skills/agentsmd-scaffold .cursor/skills/agentsmd-scaffold && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "agentsmd-scaffold" agent skill from https://github.com/majiayu000/spellbook/tree/main/skills/agentsmd-scaffold into .cursor/skills/agentsmd-scaffold/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "agentsmd-scaffold", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/majiayu000/spellbook.git --path skills/agentsmd-scaffold--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add majiayu000/spellbook --skill agentsmd-scaffold -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install majiayu000/spellbook agentsmd-scaffold --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/majiayu000/spellbook.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/skills/agentsmd-scaffold .gemini/skills/agentsmd-scaffold && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "agentsmd-scaffold" agent skill from https://github.com/majiayu000/spellbook/tree/main/skills/agentsmd-scaffold into .gemini/skills/agentsmd-scaffold/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "agentsmd-scaffold", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install majiayu000/spellbook agentsmd-scaffoldInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add majiayu000/spellbook --skill agentsmd-scaffold -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/majiayu000/spellbook.git skills-src && mkdir -p .github/skills && cp -r skills-src/skills/agentsmd-scaffold .github/skills/agentsmd-scaffold && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "agentsmd-scaffold" agent skill from https://github.com/majiayu000/spellbook/tree/main/skills/agentsmd-scaffold into .github/skills/agentsmd-scaffold/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "agentsmd-scaffold", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add majiayu000/spellbook --skill agentsmd-scaffold -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install majiayu000/spellbook agentsmd-scaffold --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/majiayu000/spellbook.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/skills/agentsmd-scaffold .opencode/skills/agentsmd-scaffold && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "agentsmd-scaffold" agent skill from https://github.com/majiayu000/spellbook/tree/main/skills/agentsmd-scaffold into .opencode/skills/agentsmd-scaffold/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "agentsmd-scaffold", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
- Skill name
agentsmd-scaffold- GitHub stars
- 287
- Token cost
- ~1.5k tokens
- SKILL.md length
- 674 words
- Files
- 4 (incl. scripts, references)
- Skills in repo
- 97
- Repo updated
- First seen
- Licence
- MIT
At a glance
Scans a repository for real evidence and proposes, or on request writes, a small stack of root and scoped AGENTS.md files with validation commands and generated-file boundaries.
- Works in 4 steps: Discover Existing Context → Choose The Instruction Stack → Produce A Candidate Plan → …
- Creating a root AGENTS.md for a repository that has none
- SKILL.md covers Operating Contract, Workflow, Decision Gates and Gotchas, plus 2 more sections
- Runs Python scripts from its folder; calls python3
What it does
The result is a plan by default, exact proposed file contents on request, and applied files only when you explicitly ask for them to be written. The agent runs scripts/scan_repo_context.py, then reads existing AGENTS.md, CLAUDE.md, WARP.md and Copilot instructions along with the README, CONTRIBUTING, manifests, Makefiles, CI workflows, generated files and risky areas such as migrations, auth, secrets and payments.
It picks the smallest instruction stack that changes agent behavior: a root file for repo-wide routing and validation, and nested files only where a directory's rules differ. Guessed commands, ownership or generated-file rules are challenged unless the repo supports them, existing files get short pointers instead of rewrites, and batch changes across several repositories need your confirmation. A reference file and an evals file support the workflow, and a separate audit skill answers whether current agent context is healthy.
When your agent uses it
- Creating a root AGENTS.md for a repository that has none
- Splitting instructions into scoped files for directories with different rules
- Improving an AGENTS.md so it lists real validation commands and generated-file boundaries
Example prompts
- “Scan this repo and propose an AGENTS.md plan, but don't write any files yet.”
- “Add a scoped AGENTS.md for the migrations folder with the commands that actually work here.”
- “Trim our root AGENTS.md into a short router and move package-specific rules into nested files.”
Requirements
- python3 to run the repository scanner
Workflow steps
4 steps, taken from the step headings in SKILL.md.
What it can do on your machine
Read from SKILL.md and the folder at commit ed52af7. 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
Ships 1 file in scripts/ (Python), which the agent can run.
Shell commands in SKILL.md call:
python3
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
AGENTS.md Scaffold loads about 1.5k tokens when it runs, and up to ~2.9k if it reads all its reference files. Until then it costs about 94 tokens; SKILL.md has 674 words of instructions outside code blocks.
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); the scripts in this folder are not scanned.
SKILL.md
The full file from majiayu000/spellbook at commit ed52af7, republished under its MIT licence (© majiayu000). 674 words, ~1,518 tokens.
Download SKILL.mdSave it as
.claude/skills/agentsmd-scaffold/SKILL.md (or your agent's skills folder). This skill also uses 3 other files; get the full folder from GitHub.- name
- agentsmd-scaffold
- description
- Generate or update repository-specific AGENTS.md instruction files from real repo evidence. Use when asked to create, design, scaffold, split, or improve root or scoped AGENTS.md files for Codex/Claude/agent workflows, especially when a repo needs directory-specific rules, validation commands, generated-file boundaries, or a short agent onboarding router.
AGENTS.md Scaffold
Use this skill to generate a small, evidence-backed AGENTS.md stack for a
repository. The output may be a plan, exact proposed file contents, or applied
files when the user explicitly asks to write them.
This skill is for instruction scaffolding. Use repo-agent-context-audit first
when the user only asks whether the repo's agent context is healthy.
Operating Contract
Default to a scoped plan before editing. Only write or modify AGENTS.md,
CLAUDE.md, WARP.md, hooks, settings, or generated docs when the user has
explicitly asked to apply the scaffold.
Direct actions:
- Run read-only discovery, scanner commands, and repo command inspection.
- Produce a scoped
AGENTS.mdplan with evidence and validation commands. - Draft exact file contents when the user asks for proposed text.
Escalate before:
- Creating or editing high-context files when the user only asked for an audit.
- Rewriting existing
AGENTS.md,CLAUDE.md, orWARP.mdinstead of adding a short pointer or scoped complement. - Batch-normalizing multiple repositories.
Evidence-backed pushback:
- Challenge new scoped files when the directory has no distinct local rules.
- Challenge guessed commands, ownership, or generated-file rules unless repo evidence supports them.
Feedback loop:
- Promote repeated false starts into
references/scaffold-agents.md, scanner signals, or eval prompts.
Workflow
1. Discover Existing Context
Run the scanner from this skill directory when possible:
python3 scripts/scan_repo_context.py <repo-root>
python3 scripts/scan_repo_context.py <repo-root> --jsonThen inspect the files that matter:
- existing
AGENTS.md,CLAUDE.md,WARP.md,.claude/instructions.md, and.github/copilot-instructions.md README.md,CONTRIBUTING.md, package manifests, Makefiles, CI workflows, and documented test commands- generated files and their generators
- high-risk directories such as migrations, deploy scripts, auth, secrets, payments, registry metadata, generated clients, and production operations
Do not infer commands or ownership from names alone. Use scanner output as a lead, then verify with actual files.
2. Choose The Instruction Stack
Read references/scaffold-agents.md before proposing files. Choose the
smallest stack that changes agent behavior:
- root
AGENTS.mdfor repo-wide routing and validation - nested
AGENTS.mdonly where directory rules differ from root - no nested file for directories that only need ordinary README context
- no bulk normalization across multiple repos until a few examples have been manually validated
3. Produce A Candidate Plan
Before editing, report:
## Scoped AGENTS Plan
| Path | Why here | Rules to include | Validation |
|---|---|---|---|
| `AGENTS.md` | <repo evidence> | <root topics> | `<command>` |
| `<dir>/AGENTS.md` | <repo evidence> | <scoped topics> | `<command>` |
## Files To Preserve
- `<existing high-context file>` - <how it will be referenced or left alone>
## Open Facts
- <missing command or ownership fact that cannot be inferred>4. Scaffold On Request
When applying the scaffold:
- keep root files short, normally 80-150 lines
- keep nested files focused on that directory's ownership, source-of-truth rules, and validation commands
- include real commands and paths, not placeholders, unless the fact is truly missing
- preserve existing high-context files unless the user requested a rewrite
- pair every prohibition with a concrete alternative, helper, generator, or command
Show full SKILL.md (273 more words)Show less
Decision Gates
| Situation | Action |
|---|---|
Existing CLAUDE.md or WARP.md is already a good router | Add a short AGENTS.md pointer only if cross-runtime routing helps. |
| Root instruction file exceeds roughly 200 lines | Propose root router plus scoped files or references. |
| Directory has generated outputs | Add scoped rules naming source of truth and regenerate/check commands. |
| Directory has distinct safety rules | Add scoped rules with escalation boundaries. |
| Directory has ordinary implementation files only | Keep guidance in root unless conventions differ. |
| Commands cannot be verified from repo evidence | Leave an open fact instead of guessing. |
Gotchas
- Do not add nested
AGENTS.mdfiles for every directory. Add them only where local rules differ from root. - Do not guess build, test, lint, or generator commands from framework names. Cite the manifest, CI workflow, script, or docs that prove the command.
- Do not overwrite an existing
CLAUDE.md,WARP.md, orAGENTS.mdjust to normalize naming. Preserve it, point to it, or propose a split first. - Do not put long architecture explanations in root
AGENTS.md; route to references or existing docs instead.
Verification
After applying changes:
- run the repo's narrow validation command for the affected scope
- run any repo-wide registry, docs, typecheck, lint, or test command named in the new instructions when practical
- rerun
python3 scripts/scan_repo_context.py <repo-root>if using the bundled scanner to confirm scoped files are discoverable
If verification cannot run, report the exact missing precondition and the command that should be run later.
Resources
scripts/scan_repo_context.py: read-only scanner for high-context files, command hints, specs, local skills, and scopedAGENTS.mdcandidates.references/scaffold-agents.md: scaffold selection rules and templates for root, generated metadata, scripts/tools, skill libraries, and tests.evals/evals.json: lightweight prompts for future behavior checks.
© majiayu000, 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 3 other files (scripts, references) in skills/agentsmd-scaffold of majiayu000/spellbook.
- SKILL.md
- evals/evals.json
- references/scaffold-agents.md
- scripts/scan_repo_context.py
Open the folder on GitHubat commit ed52af7
Compare with similar skills
AGENTS.md Scaffold 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.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| AGENTS.md Scaffold this skillmajiayu000/spellbook | 287 | — | ~1.5k | Automated safety check: Pass | MIT | |
| Intent Layercrafter-station/skills | 112 | 1 repos | ~633 | Automated safety check: Pass | MIT | |
| AI Bomcdxgen/cdxgen | 1.1k | — | ~2.5k | Automated safety check: Pass | Apache-2.0 | |
| AI Context Workspace Initkvker/ai-context-workspace | 101 | — | ~948 | Automated safety check: Pass | MIT | |
| Add Remote Endpointnextcloud/android-library | 106 | — | ~1.1k | Automated safety check: Pass | Custom licence | |
| Scaffold Clauderagnar-pwninskjold/tech-snacks | 137 | — | ~1.7k | Automated safety check: Pass | MIT |
Similar skills
AI Bom
cdxgen/cdxgen
Generates AI-BOM, MCP inventory, AI skill inventory, and AI authorship provenance documents with cdxgen, cataloging models, inference services, Hugging Face purls, MCP servers and their…
AI Context Workspace Init
kvker/ai-context-workspace
Turns existing work units or an empty directory into an AI Context Workspace by scanning materials, generating background and conventions, and installing runtime skills.
Add Remote Endpoint
nextcloud/android-library
A skill your agent uses when adding a new remote/network endpoint (a RemoteOperation / OCSRemoteOperation) to this Nextcloud Android library, or when the user says "add an endpoint", "new remote…
Scaffold Claude
ragnar-pwninskjold/tech-snacks
A skill your agent uses when the user wants to create, draft, or scaffold a CLAUDE.md (or AGENTS.md) file for their project.
Harness Init
k1ein-chen/Harness-Starter
Initialize this project with the Harness Engineering starter template.
More from majiayu000/spellbook
All 97 skills in this repoSkill Ecosystem Doctor
majiayu000/spellbook
Audits and repairs how coding-agent Skills are owned, copied and exposed across runtimes, from canonical sources to quarantine and retirement.
Product Demo Builder
majiayu000/spellbook
Plans, produces or diagnoses evidence-backed product demo videos: script, capture plan, pacing checks and verified final media built on real product behavior.
Flowguard Task Guard
majiayu000/spellbook
Single entry point that routes long or ambiguous agent tasks, checks live state, bounds autonomous loops and leaves a resumable handoff.
npm Supply Chain Check
majiayu000/spellbook
Scans a repository, its lockfiles and node_modules for known malicious npm package versions and install-time indicators, using a read-only Python scanner.
Product Manager Toolkit
majiayu000/spellbook
Product management helpers: a RICE scoring script, an interview transcript analyzer and PRD templates for prioritizing features, synthesizing research and writing requirements.
Repo Agent Context Audit
majiayu000/spellbook
Audits a repository's agent-readable context, from AGENTS.md and CLAUDE.md to skills and PRODUCT or TECH specs, and recommends the smallest useful improvements.
Related
Categories
Questions about AGENTS.md Scaffold
What does AGENTS.md Scaffold do?
Scans a repository for real evidence and proposes, or on request writes, a small stack of root and scoped AGENTS.md files with validation commands and generated-file boundaries. The result is a plan by default, exact proposed file contents on request, and applied files only when you explicitly ask for them to be written.md and Copilot instructions along with the README, CONTRIBUTING, manifests, Makefiles, CI workflows, generated files and risky areas such as migrations, auth, secrets and payments.
When should I use AGENTS.md Scaffold?
AGENTS.md Scaffold fits situations like: creating a root AGENTS.md for a repository that has none; splitting instructions into scoped files for directories with different rules; improving an AGENTS.md so it lists real validation commands and generated-file boundaries.
How do I install AGENTS.md Scaffold in Claude Code?
Run `npx skills add majiayu000/spellbook --skill agentsmd-scaffold -a claude-code`. Or copy the skill folder (skills/agentsmd-scaffold in majiayu000/spellbook) into .claude/skills/agentsmd-scaffold in your project. Claude Code loads it when a task matches its description.
How do I install AGENTS.md Scaffold in Codex?
Run `npx skills add majiayu000/spellbook --skill agentsmd-scaffold -a codex`. Or copy the skill folder (skills/agentsmd-scaffold in majiayu000/spellbook) into .agents/skills/agentsmd-scaffold in your project. Codex loads it when a task matches its description.
Can I use AGENTS.md Scaffold 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 majiayu000/spellbook --skill agentsmd-scaffold -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/agentsmd-scaffold, .gemini/skills/agentsmd-scaffold, .github/skills/agentsmd-scaffold and .opencode/skills/agentsmd-scaffold in your project.
What does AGENTS.md Scaffold need to run?
Going by SKILL.md and its folder, AGENTS.md Scaffold needs Python for the scripts in its folder and the command-line tools its instructions call (python3). Our summary lists: python3 to run the repository scanner.
Does AGENTS.md Scaffold 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 AGENTS.md Scaffold 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. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.
What licence does AGENTS.md Scaffold use?
AGENTS.md Scaffold is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
How many tokens does AGENTS.md Scaffold use?
About 1.5k tokens (SKILL.md is roughly 6.1k 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 1.4k tokens, read only when the agent opens those files.
What are the alternatives to AGENTS.md Scaffold?
Skills that share tags, products or a category with AGENTS.md Scaffold: Intent Layer (crafter-station/skills, 112 stars), AI Bom (cdxgen/cdxgen, 1.1k stars), AI Context Workspace Init (kvker/ai-context-workspace, 101 stars) and Add Remote Endpoint (nextcloud/android-library, 106 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains AGENTS.md Scaffold?
majiayu000 (a GitHub user) maintains it in majiayu000/spellbook, which has 287 GitHub stars. The repository holds 97 skills in this directory. The repository was last updated on October 8, 2026.
Source: majiayu000/spellbook on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.