The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill instructs the agent to read the .env.local file to identify project parameters. Because this file typically contains sensitive credentials, API keys, and environment secrets, its entire content is exposed to the agent context.\n- [COMMAND_EXECUTION]: The skill utilizes high-privilege tools including execute_sql and apply_migration on production databases. These tools allow for arbitrary data modification and schema changes on live systems.\n- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by ingesting and acting upon untrusted data from external platforms and the web.\n
Ingestion points: Data is pulled from Sentry via search_issues and analyze_issue_with_seer, from Supabase via get_logs, and from the open web via firecrawl_search.\n
Boundary markers: Absent; there are no instructions to use delimiters or ignore embedded instructions within the ingested data.\n
Capability inventory: The skill possesses the ability to execute arbitrary SQL on production databases, modify project source code, and update the status of external service issues.\n
Sanitization: Absent; no validation or filtering is specified for external data before it is used to plan and implement code or database changes.

iterate-post-launch