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/muthuishere/mcp-server-bash-sdk/claude-mdgit clone --depth 1 https://github.com/muthuishere/mcp-server-bash-sdkWrote 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/instructions/muthuishere/mcp-server-bash-sdk/claude-md)<a href="https://agentmods.dev/instructions/muthuishere/mcp-server-bash-sdk/claude-md"><img src="https://agentmods.dev/badge/instructions/muthuishere/mcp-server-bash-sdk/claude-md.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 | $0.02713 | $0.02713 |
| Opus 5 | $0.01357 | $0.01357 |
| Sonnet 5 | $0.00543 | $0.00543 |
| Haiku 4.5 | $0.00271 | $0.00271 |
Grade A, and why
mcp-server-bash-sdk CLAUDE.md scanned grade A 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 5d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
./test_http_transport.sh # 19 HTTP tests (drives a real server over curl) How it starts
The opening of the file, as written. The whole thing — 169 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
What this is
A Model Context Protocol server SDK in pure Bash, implementing MCP revision 2026-07-28 over both standard transports. Three layers, split by file:
mcpserver_core.sh— the protocol layer. JSON-RPC framing, method dispatch, per-request version negotiation, result envelopes, error codes. It defines functions only; sourcing it starts nothing.run_mcp_serveris the stdio entry point.mcpserver_http.sh— the Streamable HTTP binding. Parses HTTP, enforces the header↔body contract, maps errors onto status codes, then calls the sameprocess_request. It contains no protocol logic, which is what keeps the two transports from drifting.moviemcpserver.sh,examples/gitserver.sh— implementations. Set the config variables, source the core (and optionally the HTTP layer), definetool_*functions, dispatch on--http.
Users copy an implementation, not the core. examples/README.md is the walkthrough.
Commands
./test_mcpserver_core.sh # 31 unit tests, one per spec scenario
./test_mcpserver_core.sh test_discover_shape # a single test by function name
./test_conformance.sh # validate live output against the official schema
./test_http_transport.sh # 19 HTTP tests (drives a real server over curl)
./test_http_transport.sh test_bad_origin_is_403 # a single HTTP test
./docs/spikes/spike-01-stdio-viability.sh # re-measure the stdio numbers
./docs/spikes/spike-02-http-listener.sh # re-check listener/portability assumptions
./scripts/test-linux.sh all # run both suites on Debian, Alpine, Ubuntu (Docker)
# one-shot smoke test (note: every request MUST carry _meta protocolVersion)
echo '{"jsonrpc":"2.0","id":1,"method":"server/discover","params":{"_meta":{"io.modelcontextprotocol/protocolVersion":"2026-07-28","io.modelcontextprotocol/clientCapabilities":{}}}}' | ./moviemcpserver.sh
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.
- 5d ago First seen · 169 lines · 2,713 tokens per session scan A a1e4ef6a1840
mcp-server-bash-sdk CLAUDE.md is an instructions file published in the GitHub repository muthuishere/mcp-server-bash-sdk (515 stars, last pushed 1mo ago), licensed MIT. It adds 2,713 tokens to every session, about $0.0136 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
tesseron AGENTS.md
AGENTS.md instructions for Eigenwise/tesseron, covering agent guide for the tesseron monorepo, package manager, workspace layout, plugin manifest is version-coupled and releases.
tesseron CLAUDE.md
Claude Code instructions for Eigenwise/tesseron, a project described as: Expose typed web-app actions to MCP-compatible AI agents over WebSocket. TypeScript SDK + matching MCP gateway.
kit CLAUDE.md
Claude Code instructions for toreva/kit, covering kernel loop invariant — read before escalating any decision, the rule, raci — accountability is non-delegable, what this is not and if you were going to ask paul.
kit AGENTS.md
AGENTS.md instructions for toreva/kit, covering agent instructions for toreva/kit, repository purpose, guardrails, canonical sentence and session protocol.
mcp-annotated-java-sdk AGENTS.md
AGENTS.md instructions for thought2code/mcp-annotated-java-sdk, covering git commit rules and commit message format.
openrouter-mcp-multimodal AGENTS.md
AGENTS.md instructions for stabgan/openrouter-mcp-multimodal, covering agent instructions, before you ship, releasing (read this before publishing), short version and version files (must all match package.json).