Borrowing it
Nothing to install: this file belongs to sbroenne/mcp-server-excel. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/sbroenne/mcp-server-excel/main/AGENTS.mdgit clone --depth 1 https://github.com/sbroenne/mcp-server-excelWrote 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/sbroenne/mcp-server-excel/agents-md)<a href="https://agentmods.dev/instructions/sbroenne/mcp-server-excel/agents-md"><img src="https://agentmods.dev/badge/instructions/sbroenne/mcp-server-excel/agents-md/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/instructions/sbroenne/mcp-server-excel/agents-md"><img src="https://agentmods.dev/badge/instructions/sbroenne/mcp-server-excel/agents-md.svg" alt="Reviewed on agentmods" width="80" 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 | $0.03522 | $0.03522 |
| Opus 5.5 | $0.01409 | $0.01409 |
| Sonnet 5.5 | $0.00704 | $0.00704 |
| Haiku 4.5 | $0.00352 | $0.00352 |
Grade A, and why
mcp-server-excel AGENTS.md 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 3d 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 — 265 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ExcelMcp agent instructions
ExcelMcp automates installed desktop Excel through COM on Windows. Use
PowerShell 7.4 or later (pwsh) and the SDK selected by global.json. Desktop Excel is required
for COM tests; GitHub-hosted runners do not have Excel.
MCP Server and excelcli are equal entry points: behavior, defaults, validation,
results, and documentation must agree.
ExcelMcp always operates in the current user's Windows session and security context. Discover and reconnect only to Excel instances accessible in that context; do not seek access to other users' sessions or elevate permissions. Workbook location does not change this scope. Running in the user's context does not authorize changes to unrelated workbooks.
Code Review Rules
Report high-confidence defects introduced by the change, not style or unrelated cleanup. Establish a reachable failure, its affected caller/workbook state, and the missing or misleading test evidence. Review does not require builds, dependency installation, or edits unless requested.
- Trace changed catches and fallbacks through their caller. Logging an error
then returning empty/default data, skipping a failed item, or reporting success
still hides failure. Optional COM reads may suppress only exact documented
unavailable-member HRESULTs;
0x800A03ECis not a missing-property diagnosis. Unexpected COM errors must fail with their context. - Check failure ownership: Core leaves execution errors to batch/Service; transport boundaries may convert failures, not erase them. Preserve original cause, cancellation, timeout, session recovery, and exact partial state. Cancellation is not rollback; reject unrequested workbook cleanup.
- Review COM ownership across every exit and failure path. Acquired references
need
finallyrelease, not session-owned objects. Process cleanup needs PID/start-time identity and retained handles. EXL011 and EXL005 reject dynamic declarations and late-bound calls in production and test builds. - Follow changed source contracts through generated Service, CLI options/batch JSON, MCP schemas, and manual exceptions. Names, defaults, validation, results, and timeouts must agree. Adapters must not acquire Core/session ownership.
- Check MCP protocol output, including bootstrap paths, not just direct console calls. Telemetry must never wait on setup, export, flush, or disposal on product paths, delay incremental results/cancellation/exit, or alter results/errors. Bound queues; verify local delivery separately from responsiveness and privacy.
- Check guidance against implemented inputs, defaults, and capabilities. Distinguish MCP snake_case inputs, CLI flags, nested JSON keys, and outputs. Flag nonexistent actions, forced unrequested work, emojis, required unattended screenshots, and consent prose presented as implemented server elicitation.
- Tests must establish returned fields and actual Excel state, including the partial state after failure; success alone or mocked COM is not evidence. Compiler checks catch supported patterns, not correct HRESULT semantics, complete COM ownership, runtime state, or matching entry-point behavior.
- Use
ponytail-reviewwhen available after correctness/security review for concrete behavior-preserving simplifications. Required COM cleanup, validation, failure reporting, tests, and generated contracts are not bloat.
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.
- 3d ago Changed · +10 lines · +140 tokens per session 6eab6ab935e8
- 7d ago Changed · +3 lines · +41 tokens per session 61b4d04f0ab3
- 12d ago Changed · +18 lines · +251 tokens per session c9bac530fc5a
- 14d ago Changed · +3 lines · +43 tokens per session a81001cfe855
- 17d ago Changed · +17 lines · +222 tokens per session d44f77eecb76
- 19d ago Changed · +3 lines · +48 tokens per session 5dab65f50401
- 21d ago Changed · +13 lines · +152 tokens per session c365dc122bdf
- 22d ago Changed · -6 tokens per session e0dcab62d0fd
- 23d ago Changed · +15 lines · +207 tokens per session 727a5b0ef990
- 25d ago First seen · 183 lines · 2,424 tokens per session scan A daf18c519335
mcp-server-excel AGENTS.md is an instructions file published in the GitHub repository sbroenne/mcp-server-excel (853 stars, last pushed yesterday), licensed MIT. It adds 3,522 tokens to every session, about $0.0141 per session on Opus 5.5. 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 2026-08-30.
mcp-server-excel AGENTS.md compared with instructions from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| mcp-server-excel AGENTS.mdsbroenne/mcp-server-excel | Instructions file | 853 | yesterday | MIT | A | 3,522 | Codex, OpenCode |
| vscode buildNext.instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 9,719 | GitHub Copilot |
| next.js AGENTS.mdvercel/next.js | Instructions file | 143k | today | MIT | A | 7,461 | Codex, OpenCode |
| langchain AGENTS.mdlangchain-ai/langchain | Instructions file | 149k | yesterday | MIT | A | 4,680 | Codex, OpenCode |
| vscode oss-third-party-notices.instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 5,516 | GitHub Copilot |
| vscode copilot-instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 3,093 | GitHub Copilot |
| deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness | Instructions file | 274k | 2d ago | MIT | A | 4,313 | Codex, OpenCode |