The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from transcript files and project codebase context to generate reports.
Ingestion points: Processes external transcript files provided by the user and reads codebase files (e.g., AGENTS.md, CLAUDE.md, commit history) to resolve domain terms.
Boundary markers: No explicit instructions or markers are used to prevent the agent from following instructions embedded within the transcript text.
Capability inventory: The skill has permissions to read files across the local filesystem (Step 2), write an HTML file (Step 5), and open the default browser.
Sanitization: The provided template.html uses the v-html directive in Vue.js to render summaries, notes, and decisions. This allows raw HTML from the transcript analysis to be executed in the browser context without sanitization, leading to a potential XSS vulnerability if the transcript contains malicious payloads.
[COMMAND_EXECUTION]: The skill performs filesystem operations and initiates browser execution.
File Access: Reads multiple files across the current and related repositories to gather context.
Output Generation: Writes a new file call-report.html to the disk.
Execution: Instructs the agent to open the generated report in the system's default browser using shell-level commands.

call-report