calphonse-skills
AGENTS.md1 starsChanged 10 days agoMITView on GitHub
All skills live at the repository root as {skill-name}/SKILL.md. AGENTS.md is the canonical index — keep it in sync with the filesystem. Install any skill with the Skills CLI:
Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.
Of the AGENTS.md files we read, the 31,774 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
In this repository29 other files it gives its agents.
- vercel-react-best-practices/AGENTS.mdAGENTS.md
- application-lifecycle-traceSkill
- architecture-reviewSkill
- awsSkill
- codebase-mapperSkill
- code-review-frontendSkill
- comment-sickoSkill
- cspSkill
- demo-scriptSkill
- documentation-writerSkill
- dotnetSkill
- feature-orchestratorSkill
- frontend-patternsSkill
- lets-talk-about-itSkill
- nextjs-proxy-csp-hardeningSkill
- owasp-securitySkill
- pr-body-styleSkill
- prd-modeSkill
- prove-it-worksSkill
- security-reviewSkill
- simplifySkill
- sweSkill
- systems-architectureSkill
- task-generation-modeSkill
- thermo-nuclear-code-quality-reviewSkill
- typescript-react-reviewerSkill
- typescript-strict-styleSkill
- vercel-react-best-practicesSkill
- whats-going-onSkill
The file
We split this AGENTS.md into the rules it gives an agent
5 rules. About 0 are common, 2 are distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 3
All skills live at the repository root as {skill-name}/SKILL.md.
- Distinctive: line 48
Prefer referencing another skill in this repo over duplicating its content.
- Very specific: line 49
If a skill links to references/*.md, commit those files.
- Distinctive: line 50
Remove references to tools, folders, or sibling skills that no longer exist.
- Very specific: line 51
No URLs or real hostnames in any file here.
≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Skills Index
All skills live at the repository root as {skill-name}/SKILL.md. AGENTS.md is the canonical index — keep it in sync with the filesystem.
Install any skill with the Skills CLI:
npx skills add ChristopherAlphonse/calphonse-skills --skill <name>Skill Index
| Skill | Description |
| application-lifecycle-trace | Trace and explain an application's lifecycle end to end |
| architecture-review | Read-only architecture and distributed-systems review — complexity, bottlenecks, reliability risks, and cost savings, with findings marked VERIFIED / [ASSUMPTION] / [NEEDS MEASUREMENT] |
| aws | Index and router for AWS work — CLI/SDK, IaC, IAM, networking, storage, databases, compute, containers, serverless, messaging, CI/CD, security, observability, billing, Bedrock/SageMaker, Cognito auth |
| code-review-frontend | Multi-agent PR review orchestrator — discovers quality tooling, spawns parallel specialist agents, reports verified findings only, never fixes |
| codebase-mapper | Structured onboarding map of an existing codebase — tech stack, architecture, key flows, conventions, known concerns |
| comment-sicko | A deranged comment-hater that savors deletion and condemns workaround code |
| csp | Advanced CSP bypass techniques — policy weaknesses, nonce leakage, trusted-endpoint abuse, exfiltration channels CSP can't block |
| demo-script | Short, natural, spoken demo script that explains what changed, shows it working, and ties it to business value |
| documentation-writer | Diátaxis documentation expert — tutorials, how-tos, reference, explanation — plus product copy principles aligned with brand voice |
| dotnet | Index and router for C#/.NET work — ASP.NET Core Web API endpoints, DTOs, error handling middleware, OpenAPI/Swagger, and design pattern reviews (Command, Factory, Repository, Provider, DI, SOLID) |
| feature-orchestrator | Session 1 of a two-session feature workflow — settle a feature in conversation, then hand off a copy-ready orchestration prompt with a chosen sub-agent model roster |
| frontend-patterns | Frontend development patterns for React, Next.js, state management, performance optimization, and UI best practices |
| lets-talk-about-it | One-pass interrogation plus coding guardrails — explicit assumptions, simplicity, surgical edits, verifiable success criteria |
| nextjs-proxy-csp-hardening | Build and review secure Next.js proxy.ts files — CSP, nonce-based script protection, browser security headers, middleware.ts migration |
| owasp-security | OWASP Top 10 secure coding guidance — auth, XSS, SQLi, CSRF, API security |
| pr-body-style | Draft, rewrite, or audit GitHub PR bodies — accurate, short, business-focused |
| prd-mode | Turn abstract ideas into actionable project specs that align engineering and design on the problem and success metrics |
| prove-it-works | Verify a finished task against the real artifact, not a proxy, self-report, or "it compiles", then write a UAT document so the user can confirm it themselves |
| security-review | Scoped security checklist and patterns for auth, user input, secrets, API endpoints, and payment flows |
| simplify | Restate the last message in plain human language, with no jargon |
| swe | Lazy senior dev mode — smallest change that fully solves the task, and a reply a busy human reads once |
| systems-architecture | Generates a high-level system architecture document from lifecycle documentation — actors, interactions, technologies |
| task-generation-mode | Converts a PRD into a structured, actionable implementation task list with verifiable outcomes |
| thermo-nuclear-code-quality-review | Extremely strict maintainability review — abstraction quality, giant files, spaghetti-condition growth |
| typescript-react-reviewer | Expert code reviewer for TypeScript + React 19 applications — anti-patterns, state management, type safety |
| typescript-strict-style | Strict TS/JS discipline — assertion safety, exhaustive error handling, function size, naming, performance, dependency minimalism |
| vercel-react-best-practices | React and Next.js performance optimization guidelines from Vercel Engineering |
| whats-going-on | One-pass codebase intelligence sweep — core use cases, lifecycle, architecture, and where to start reading |
Conventions
- Front-matter
name:should match the folder name. - Durable output belongs under
.planning/*in the target project, never in this repo. - Prefer referencing another skill in this repo over duplicating its content.
- If a skill links to
references/*.md, commit those files. A link to a missing file silently disables the workflow that depends on it. - Remove references to tools, folders, or sibling skills that no longer exist.
- No URLs or real hostnames in any file here. Never fetch, browse, or search the web while using these skills unless the user explicitly asks in that conversation. Use placeholders like
API_ORIGINin examples.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.