The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/autocorrect.ts uses the execSync method to run shell commands.
Evidence: It executes npx autocorrect-node --fix on a file path, which could potentially be exploited if the filename contains shell metacharacters.
[EXTERNAL_DOWNLOADS]: The skill relies on runtime downloads of packages to perform its tasks.
Evidence: SKILL.md instructs the agent to use npx -y bun to run the main logic, and scripts/autocorrect.ts uses npx to execute the autocorrect-node tool from the NPM registry.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it analyzes and summarizes user-provided content without sanitization.
Ingestion points: Reads user-provided markdown or plain text files in Step 1.
Boundary markers: None; the agent is not instructed to ignore formatting instructions inside the source file.
Capability inventory: The skill can write to the filesystem (writeFileSync in scripts/main.ts) and execute shell commands (execSync in scripts/autocorrect.ts).
Sanitization: No sanitization or escaping is performed on the input content before generating titles or summaries.

baoyu-format-markdown