The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/autocorrect.ts executes external shell commands using spawnSync to run npx autocorrect-node --fix [filePath]. This allows for the execution of a third-party tool on the user's filesystem.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it processes untrusted user-provided content during its 'Analyze' phase (Step 2 in SKILL.md). Malicious instructions embedded in a markdown file could influence agent behavior.
Ingestion points: The agent reads a user-specified file path in Step 1 of the workflow.
Boundary markers: None; the skill does not use delimiters or instructions to ignore embedded commands within the processed file content.
Capability inventory: Across all scripts, the skill can perform filesystem writes (writeFileSync in scripts/main.ts) and shell command execution (spawnSync in scripts/autocorrect.ts).
Sanitization: There is no evidence of sanitization or filtering of external content before it is interpolated into the agent's context for analysis.
[EXTERNAL_DOWNLOADS]: The typography process relies on npx to run autocorrect-node. If the package is not cached locally, it will be downloaded from the NPM registry at runtime.

baoyu-format-markdown