Borrowing it
Nothing to install: this file belongs to morluto/jacobian. 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/morluto/jacobian/main/.agents/skills/audit-mcp-tool-friction/SKILL.mdgit clone --depth 1 https://github.com/morluto/jacobianWrote 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/skills/morluto/jacobian/audit-mcp-tool-friction)<a href="https://agentmods.dev/skills/morluto/jacobian/audit-mcp-tool-friction"><img src="https://agentmods.dev/badge/skills/morluto/jacobian/audit-mcp-tool-friction/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/skills/morluto/jacobian/audit-mcp-tool-friction"><img src="https://agentmods.dev/badge/skills/morluto/jacobian/audit-mcp-tool-friction.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 2 findings, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high System Prompt Leakage · line 43 Skill contains instructions that could directly expose system prompts, internal rules, or hidden instructions to users or external parties.Fix: Remove any instructions that reveal, print, or output system prompts or internal rules. System instructions should never be exposed to end users.
- medium Excessive Agency · line 23 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
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 | $0.00026 | $0.00882 |
| Opus 5 | $0.00013 | $0.00441 |
| Sonnet 5 | $0.00005 | $0.00176 |
| Haiku 4.5 | $0.00003 | $0.00088 |
Grade A, and why
audit-mcp-tool-friction 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 — 92 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Audit MCP Tool Friction
Audit whether an agent that has chosen a tool can form a valid request, recover from a failure, and use the returned value. This is a public-contract audit, not a call-count target and not a test of model willingness to use a tool.
Set the boundary
Record the tool, revision, client, task or trace, and public inputs inspected. Keep these questions separate:
- Adoption: whether the agent saw, discovered, or selected a tool.
- Friction: whether the selected tool made its contract, valid input, error recovery, and result use clear enough to execute correctly.
- Capability: whether an operation actually covers the requested work.
Do not call a successful Python fallback evidence that a tool contract is easy to use. Do not call a single malformed request a product defect without checking what the public contract exposed beforehand.
Inspect the public contract
For each relevant operation, inspect the generated MCP input and output schemas, tool description, discovery/inspection result, operation examples, and error messages. Trace every non-obvious requirement to its owner:
| Contract element | Check |
|---|---|
| Input expression | Required values, units, encodings, ordering, and bounds are visible before the call. |
| Schema projection | Constraints representable in JSON Schema appear there with useful field descriptions and examples. |
| Semantic validation | Rules enforced only by a model validator have an explicit description and minimal valid example. |
| Recovery | Errors identify the field and state the smallest concrete correction. |
| Result use | Result shape, exactness, limits, and reconstruction information let the caller continue without guessing. |
Pydantic does not infer JSON-Schema descriptions or examples from custom model
validators. Add Field(description=..., examples=...) or appropriate
json_schema_extra when a semantic invariant cannot be encoded as a standard
JSON-Schema constraint. Keep canonical output rules when they support stable
composition; improve the input guidance before weakening the invariant.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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 Changed · -57 tokens per session 532db849fc3e
- 11d ago First seen · 92 lines · 83 tokens per session scan A 28c96c1e803d
audit-mcp-tool-friction is a skill published in the GitHub repository morluto/jacobian (191 stars, last pushed yesterday), licensed MIT. It adds 26 tokens to every session and 882 once invoked, about $0.0001 per session on Opus 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.
Other skills, from other repositories
witness
Independently witness that an Allium loop's convergence claim is true and was reached honestly. Use when the user wants to verify a loop's self-report, confirm tests really pass and no generated test was weakened, produce a convergence certificate or witness record, gate CI on a trustworthy signal, or check that an…
trace-mcp-refactoring
Safe refactoring workflow using trace-mcp — assess risk, find candidates, check impact, and rename symbols across all files without missing import sites or cross-file references.
m1nd-first
Use when investigating a repository, searching for implementation, reviewing changes, working from specs/docs, or preparing a risky code change in an environment where m1nd is available. This doctrine makes m1nd the first investigative layer before grep, glob, or manual file reads, except when the task is pure…
fsl-requirements-document
Generate, edit, and re-verify a human-readable requirements document (Markdown) from a checked FSL requirements/spec dialect file, using fslc document generate/claims/check. The agent is a non-normative editor and review-support assistant, never a compiler — it may only fill in the document's editable slot…
review-thread
Record review verdicts and run the StateNote review thread on an agent-operation instance — every ReviewRequest verdict is one atomic batch-direct-write that advances status and co-writes a reviewnote explaining why, and findings are corrected by appending notes rather than editing them.
audit-architecture
Audit module coupling, cohesion, hidden dependencies, and design smells. Use when conducting architecture reviews, evaluating design decisions, or identifying structural tech debt.