The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local shell script, scripts/estimate-tokens.sh, and several ecosystem-standard audit tools including npm audit, pip-audit, govulncheck, and cargo audit to analyze the project environment.\n- [EXTERNAL_DOWNLOADS]: The skill uses npx to run utilities such as better-npm-audit, license-checker, and depcheck, which involves fetching software packages from the public NPM registry.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its design, which involves ingesting entire, potentially untrusted codebases into the model's context for analysis.\n
Ingestion points: The skill uses Read and Glob tools to load source files, configurations, and entry points as specified in references/report-structure.md and SKILL.md.\n
Boundary markers: Absent. The documentation does not specify the use of delimiters, XML tags, or explicit instructions to disregard commands embedded within the audited source files.\n
Capability inventory: The agent retains access to powerful tools including Bash, Grep, Read, and Task management while the untrusted code content is present in its context.\n
Sanitization: Absent. There is no evidence of logic for sanitizing or escaping project file contents before they are processed by the model.

audit-full