The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources and local vault files to inform the PRD generation process.
Ingestion points: Reads existing PRD files from the vault (04-projects/*/PRDs/*.md) and fetches issue content/discussions from GitHub and Linear via their respective CLI tools and APIs.
Boundary markers: The instructions do not define explicit delimiters or 'ignore' instructions for the data gathered from external trackers.
Capability inventory: The skill has the ability to execute shell commands (bash), write files to the local vault, and perform network requests to publish content to Confluence, Notion, and HackMD.
Sanitization: There is no explicit evidence of sanitization or filtering of the content retrieved from external issue trackers before it is processed by the agent.
[COMMAND_EXECUTION]: The skill executes shell commands via Bash for utility purposes.
Evidence: Uses date to generate timestamps and mkdir -p to ensure directory structures exist in the local vault.
Evidence: Interacts with the GitHub CLI (gh search issues) to retrieve context, which interpolates user-provided keywords into the command string.
[EXTERNAL_DOWNLOADS]: The skill interacts with several well-known external services to gather context and publish documents.
Evidence: Fetches issue data from GitHub and Linear.
Evidence: Supports publishing documents to Confluence, Notion, and HackMD via their official APIs or integration tools.

generate-prd