The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the session_store database which can lead to indirect prompt injection attacks.
Ingestion points: Reads conversation history from the turns table in the SQLite database (identified in SKILL.md).
Boundary markers: None identified. The instructions do not define delimiters or provide guidance to ignore instructions embedded within retrieved session turns.
Capability inventory: Uses the sql tool to query databases and the copilot CLI tool to resume execution environments.
Sanitization: No sanitization or filtering is applied to the retrieved user_message or summary fields before they are presented to the agent.
[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands to resume sessions and list repository data.
Evidence: Execution of copilot --resume SESSION_ID and gh issue list based on data retrieved from the database.

session-recovery