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.
git clone --depth 1 https://github.com/KaelSensei/MagicAIBuilderWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/kaelsensei/magicaibuilder/search-docs)<a href="https://agentmods.dev/commands/kaelsensei/magicaibuilder/search-docs"><img src="https://agentmods.dev/badge/commands/kaelsensei/magicaibuilder/search-docs.svg" alt="Measured on agentmods" height="20"></a>What 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 5.1 | $0.00000 | $0.01544 |
| Opus 5 | $0.00000 | $0.00772 |
| Sonnet 5 | $0.00000 | $0.00309 |
| Haiku 4.5 | $0.00000 | $0.00154 |
Grade A, and why
search-docs scanned grade A with 0 findings 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 6d 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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 126 lines — stays where its author put it; the contents beside it link to each section on GitHub.
search-docs Command – Search Project Documentation via MCP
When /search-docs <query> is invoked, automatically search the project documentation using the MCP server configured in .cursor/mcp.json (e.g. the context7 server) and present the most relevant excerpts and links.
Step 1: Load Project Context & Rules
- Assume the project root as the working directory
- Load and strictly follow ALL Cursor rules from
.cursor/rules/*.mdc:security.mdc– Security requirements (no leaking secrets, tokens, or private data)technical-stack.mdc– React Native/SQLite and tooling conventionsdocumentation.mdc– Documentation update requirementsversion-management.mdc– Git commit/push workflowgeneral-principles.mdc– Project philosophy and quality bar
- Read high‑level docs to understand the documentation layout:
README.mddocs/(especiallydocs/installation/,docs/architecture/,docs/development/,docs/troubleshooting/)PROGRESS.mdfor current feature set
- Read
.cursor/mcp.jsonto discover available MCP servers (e.g."context7") and their URLs/headers - Identify the current Git branch and assume it is a utility/feature branch, not
main
Step 2: Parse the Search Request
- Parse the natural‑language query after
/search-docs(e.g./search-docs how to build release apk) - Normalize and expand the query with project context when useful, for example:
- Map “release build” →
ANDROID_CONFIG,PLAY_STORE_DEPLOYMENT,RUN_ON_ANDROID - Map “life counter” →
docs/feature/feature-mtg-life-counter.md - Map “local deck builder” →
docs/feature/feature-local-deck-builder.md
- Map “release build” →
- If the query is ambiguous (e.g. “install”), briefly infer likely intents (Android, iOS, dev env) instead of asking the user to rephrase; you can run multiple focused searches
Step 3: Discover & Use MCP Documentation Tools
- Use the MCP configuration from
.cursor/mcp.json(e.g. thecontext7HTTP server) and follow its own docs:- Always read the MCP tool descriptors in
mcps/<server>/tools/*.jsonbefore calling a tool - Respect any rate limits or usage guidelines documented there
- Always read the MCP tool descriptors in
- Prefer semantic / full‑text search tools exposed by the MCP server for documentation lookups, for example:
searchDocsor similar tools that indexdocs/**,README.md,PROGRESS.md, etc.- If there is a dedicated “documentation” or “knowledge base” tool, use it first
- When no suitable MCP search tool exists, fall back to local codebase search tools:
Grepfor keyword search acrossdocs/**and.mdfilesSemanticSearchwhen you need concept‑level matches (e.g. “how does local deck builder persist data?”)
- Never send secrets, environment files, or private user data to external MCP servers
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.
- 6d ago First seen · 126 lines · 0 tokens per session scan A 3b41a1757595
search-docs is a command published in the GitHub repository KaelSensei/MagicAIBuilder (2 stars, last pushed 4d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,544 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
add-endpoint
Scaffold a new API endpoint with full layering.
review
Review code changes against project conventions.
fix-issue
Investigate and fix an issue.
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.