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 instructions/agenticpawan/fullstack-pilot/claude-mdgit clone --depth 1 https://github.com/AgenticPawan/FullStack-PilotWhat 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 | $0.04862 | $0.04862 |
| Opus 5 | $0.02431 | $0.02431 |
| Sonnet 5 | $0.00972 | $0.00972 |
| Haiku 4.5 | $0.00486 | $0.00486 |
Grade C, and why
FullStack-Pilot CLAUDE.md scanned grade C with 1 finding 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 2d ago.
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.
Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
- `bash-guard.js` — PreToolUse/Bash: blocks git force-push, hard-reset, DROP TABLE without WHERE, Azure deployments outside branch; warns on wide `rm -rf` and prod builds How it starts
The opening of the file, as written. The whole thing — 255 lines — stays where its author put it; the contents beside it link to each section on GitHub.
FullStack Pilot — Developer Guide
Purpose
FullStack Pilot is a Claude Code plugin marketplace for full-stack Microsoft shops.
It ships codebase-governance skills, agents, and hooks for Angular, .NET, SQL Server,
and Azure projects. Hosted publicly at AgenticPawan/FullStack-Pilot; installed via:
/plugin marketplace add AgenticPawan/FullStack-Pilot
Repository layout
.claude-plugin/marketplace.json ← marketplace catalog (required fields: name, owner, plugins)
plugins/
pilot-core/ ← shared governance utilities
pilot-angular/ ← Angular / TypeScript rules
pilot-dotnet/ ← C# / ASP.NET Core rules
pilot-sql/ ← SQL Server / EF Core rules
pilot-azure/ ← Azure / Bicep / ACA rules
pilot-rag/ ← self-hosted RAG scaffold (/fsp-rag-init)
scripts/validate.mjs ← zero-dependency CI validator
.github/workflows/validate.yml ← runs validate.mjs on every push/PR
Each plugin directory MUST have:
<plugin>/
.claude-plugin/plugin.json ← manifest: name, version, description, author
commands/fsp-<verb>.md ← commands (legacy — prefer SKILL.md with name: fsp-<verb>)
skills/<skill-name>/SKILL.md ← skills (when added)
agents/<name>.md ← agents (when added)
hooks/hooks.json ← hooks (when added)
Command conventions
- Command files MUST be named
fsp-<verb>.md(invoked as/fsp-<verb>). Thefsp-prefix brands every FullStack Pilot command; CI enforces it. - Legacy pattern:
commands/fsp-<verb>.mdfiles are still supported but new commands should be authored as aSKILL.mdwithname: fsp-<verb>in frontmatter — the Skill tool resolves both forms, and SKILL.md carries richer routing metadata.
Plugin manifest conventions
userConfig: declare user-overridable options inplugin.jsonas auserConfigobject (key: { title, type, default, description }). Hook scripts read the live value viaprocess.env.CLAUDE_PLUGIN_OPTION_<KEY_UPPERCASE>(boolean options arrive as'true'/'false'strings). PreferuserConfigkill-switches over hard-coded behaviour for anything ops teams may need to turn off per-project.dependencies: every stack plugin MUST declare"dependencies": [{ "name": "pilot-core" }]— CI enforces this.pilot-coreis the base and is exempt.pilot-ragalso declares this dependency so its hooks and skills load only when pilot-core's security floor is present.defaultEnabled:pilot-ragMUST declare"defaultEnabled": false— the RAG scaffold is opt-in and requires/fsp-rag-initto generate the project before it is usable. CI enforces this. Stack plugins omitdefaultEnabled(defaults totrue).
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.
- 2d ago First seen · 255 lines · 4,862 tokens per session scan C 3d80f1437caf
FullStack-Pilot CLAUDE.md is an instructions file published in the GitHub repository AgenticPawan/FullStack-Pilot (2 stars, last pushed 1mo ago), licensed MIT. It adds 4,862 tokens to every session, about $0.0243 per session on Opus 5. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
azure-sdk-for-rust changelog.instructions.md
Instructions for Azure/azure-sdk-for-rust, a project described as: This repository is for the active development of the Azure SDK for Rust. For consumers of the SDK we recommend visiting Docs.rs and looking up the docs for any of libraries in the SDK.
building-a-coding-agent-from-scratch-course AGENTS.md
Instructions for decodingai-magazine/building-a-coding-agent-from-scratch-course, covering decode, project structure, tech stack, docs & external services and running commands.
GPT-RAG release.instructions.md
Instructions for Azure/GPT-RAG, a project described as: Sharing the learning along the way we been gathering to enable Azure OpenAI at enterprise scale in a secure manner. GPT-RAG core is a Retrieval-Augmented Generation pattern running in Azure, using Azure Cognitive Search for retrieval and Azure OpenAI large…
azure-dev copilot-instructions.md
Instructions for Azure/azure-dev: For any work in this repository, especially for code reviews, you MUST read cli/azd/AGENTS.md in its entirety (every line) first.
apex-accelerator no-interactive-shell.instructions.md
Prevents interactive shell prompts and long-output terminal replays from being injected into chat. Forbids -i flags on mv/rm/cp, read -p, and confirm prompts (incl. inside bash -c '...'). Pipe long output to files. Scoped to chat-context-loaded files; skill references/ and templates/ are exempt because they hold…
dxkit CLAUDE.md
Instructions for vyuh-labs/dxkit, covering claude.md — dxkit development rules, architecture rules, 1. tool invocation goes through the registry, 2. never duplicate tool invocation logic and 3. language facts come from detect.ts.