Back to skill
Skillv1.0.1
ClawScan security
ctx-shrink · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignApr 5, 2026, 5:03 PM
- Verdict
- benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- The skill is an instruction-only wrapper for a local codebase analyzer; its requirements and instructions are coherent with its stated purpose and it does not request extra credentials or installs, but you should verify the actual ctx-shrink binary/repository before executing it on sensitive projects.
- Guidance
- This skill is coherent: it tells the agent to run a local analyzer and read its report. Before using it, confirm you trust the ctx-shrink executable/repository (the README points to a GitHub repo). Do not run unreviewed install scripts or pipe installers into a shell. Be aware that scanning a project lets the tool/agent read all files in that project (including any secrets); only run it on projects you intend to expose to the tool, and review AI-CONTEXT.md outputs before sharing them externally.
Review Dimensions
- Purpose & Capability
- okName/description (codebase analyzer, publish-safety checks) match the instructions: run a local tool (ctx-shrink) against the user's project and read its report. No unrelated credentials, binaries, or config paths are requested.
- Instruction Scope
- okSKILL.md instructs the agent to run ctx-shrink on the user's project directory and read the output file (AI-CONTEXT.md or a custom file). This requires access to the project's files, which is expected for this functionality. The instructions do not ask the agent to read unrelated system files or exfiltrate data to external endpoints.
- Install Mechanism
- noteNo install spec is included (instruction-only), which minimizes scripted-install risk. The README recommends cloning and running a checked-in script from a GitHub repo; that is a user-mediated install path — verify the repository and the script before running. The skill itself does not supply or automatically fetch binaries.
- Credentials
- okNo environment variables, credentials, or config paths are requested. This is proportionate for a local analysis tool. There are no unexpected secret-related requirements.
- Persistence & Privilege
- okalways is false and autonomous invocation is allowed only by the platform default. The skill does not request persistent installation or modify other skill/system settings. No elevated privileges are requested in the SKILL.md.