Is mcp-tool safe to install?

SaferSkills independently audited mcp-tool (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.

CategoryWeightCategory scoreContribution

Securityprompt, exec, net, exfil, eval

35%

100

35.0 pts

Supply chainhash, typosquat, maintainer, lockfile

20%

100

20.0 pts

Maintenancestaleness, pinning, CI

15%

100

15.0 pts

TransparencySKILL.md, perms, README

15%

100

15.0 pts

Communityinstalls, verify, response

15%

100

15.0 pts

Findings & checks · 0 flagged

Securityscore 100 · 0 findings

✓—All security checks passedNo findings in this category for the latest scan.pass

Supply chainscore 100 · 0 findings

✓—All supply chain checks passedNo findings in this category for the latest scan.pass

Maintenancescore 100 · 0 findings

✓—All maintenance checks passedNo findings in this category for the latest scan.pass

Transparencyscore 100 · 0 findings

✓—All transparency checks passedNo findings in this category for the latest scan.pass

Communityscore 100 · 0 findings

✓—All community checks passedNo findings in this category for the latest scan.pass

The primary manifest — the file an agent reads to learn what this artifact does.

.claude/skills/mcp-tool/SKILL.md · 1 file

.claude/skills/mcp-tool/SKILL.md1.2 KB · Markdown

/mcp-tool

Add a new MCP tool.

Steps

Add the tool to `spec/mcp-tools.yaml` (source of truth, R-meta-1):

Name: clawops_<verb> or clawops_<noun>_<verb>
Assign toolset (cli | workflow | read | admin — or array)
Write description with "Use when" and "Do NOT use when" (R3)
Set ALL four annotations hints (R10): readOnlyHint, destructiveHint, idempotentHint, openWorldHint
Define input fields with types and optional/default

Regenerate `_generated.ts`:

   pnpm gen:schemas

Create the handler in the appropriate toolset subdirectory:

   src/mcp/tools/<toolset>/<tool-name>.ts

Register the handler in src/mcp/server.ts.

Write tests in tests/mcp/tools.test.ts — invoke via in-memory MCP client and assert schema conformance.

Run `pnpm typecheck && pnpm test`.

Constraints

Per .claude/rules/mcp.md: no credentials in inputs (R6), stdio never writes stdout (R15), cap output at 8KB (R14).
Destructive tools must trigger elicitation (R19) unless yes: true.
Long-running tools (>10s) must emit notifications/progress or return a taskId.

mcp-tool— agent skill

Is mcp-tool safe to install?

/mcp-tool

Steps

Constraints

Audit the pieces. Scan the whole. Decide.