Borrowing it
Nothing to install: this file belongs to RipeMangoBox/BITE. 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/RipeMangoBox/BITE/main/.claude/skills/skill-fit-guard/SKILL.mdgit clone --depth 1 https://github.com/RipeMangoBox/BITEWrote 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/ripemangobox/bite/skill-fit-guard)<a href="https://agentmods.dev/skills/ripemangobox/bite/skill-fit-guard"><img src="https://agentmods.dev/badge/skills/ripemangobox/bite/skill-fit-guard.svg" alt="Measured on agentmods" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00056 | $0.00422 |
| Opus 5 | $0.00028 | $0.00211 |
| Sonnet 5 | $0.00011 | $0.00084 |
| Haiku 4.5 | $0.00006 | $0.00042 |
Grade A, and why
skill-fit-guard 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 8d 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.
What it actually says
Skill Fit Guard
What this skill does
This skill handles recurring skill mismatch after a .claude skill call.
It does not auto-modify any skill. It asks for user consent before any skill spec change.
Triggering conditions
Use this skill only when both conditions are true:
- A just-called skill is not aligned with user intent (scope/stage/input/output mismatch).
- The same mismatch is likely to recur (pattern-level issue, not one-off noise).
Inputs
- Invoked skill name
- User's actual intent
- Observed mismatch signals
- Recurrence evidence (if any)
Diagnosis framework
Diagnose mismatch in one or more buckets:
- Intent mismatch: chosen skill solves a different goal.
- Stage mismatch: correct capability, wrong pipeline stage.
- Input contract mismatch: required inputs differ from user-provided context.
- Output contract mismatch: user expects different granularity/format.
- Strictness mismatch: too harsh/too shallow for the requested mode.
Output contract
Always return four blocks:
- Observed mismatch: short, evidence-based summary.
- Why recurrence is likely: what in trigger/spec causes repeated misfire.
- Revision options: 2-4 options (light/medium/heavy).
- Consent question: ask whether to modify the skill now.
Option menu guidance
Offer options similar to:
- Light: tune trigger wording and non-goals.
- Medium: refine workflow and output contract.
- Heavy: split one skill into two focused skills.
- No change: keep current skill and add usage note only.
Boundaries
- No autonomous edits to skill files.
- No forced workflow switch.
- No repeated prompting after user declines.
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.
- 8d ago First seen · 62 lines · 56 tokens per session scan A 999cc5af104d
skill-fit-guard is a skill published in the GitHub repository RipeMangoBox/BITE (61 stars, last pushed 25d ago), licensed MIT. It adds 56 tokens to every session and 422 once invoked, about $0.0003 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
results-report
A workflow for turning completed experiment analyses into a structured research report with findings, limitations, failures, and next steps.
code-understand
Build a ranked, citation-backed focus map of a codebase — the files and symbols an architecture actually hangs on — using CodeGraph when available and a dependency-free builtin (regex AST extraction + rg cross-file references) otherwise. Use this any time you (the coding agent) need to orient in an unfamiliar or large…
api-errors
McpError constructor, JsonRpcErrorCode reference, and error handling patterns for @cyanheads/mcp-ts-core. Use when looking up error codes, understanding where errors should be thrown vs. caught, or using ErrorHandler.tryCatch in services.
report-issue-framework
File a bug or feature request against @cyanheads/mcp-ts-core when you hit a framework issue. Use when a builder, utility, context method, or config behaves contrary to the documented API — not for server-specific application bugs.
report-issue-local
File a bug or feature request against this MCP server's own repo. Use for server-specific issues — tool logic, service integrations, config problems, or domain bugs that aren't caused by the framework.
link-health
Use at the start of Link work when readiness is unclear, after installs or upgrades, and before repairs; verify health, inspect interrupted writes, back up, and repair generated indexes without MCP.