The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes several local commands to gather context for work summarization.
Uses curl to interact with a local API server at http://localhost:4096 to fetch session metadata, messages, and todo items.
Executes git diff, git diff --cached, and git log to analyze local repository state.
Utilizes sqlite3 to query the local database at ~/.local/share/opencode/opencode.db as a fallback mechanism.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through the processing of untrusted data.
Ingestion points: Fetches message content and titles from previous agent sessions via the /session/<SESSION_ID>/message API endpoint.
Boundary markers: Instructions do not specify explicit delimiters or "ignore" instructions when interpolating fetched session text into the summarization prompt.
Capability inventory: The agent possesses the ability to execute shell commands (curl, git, sqlite3) and read local files.
Sanitization: There is no evidence of sanitization or filtering of the historical session text before it is processed by the LLM for summarization.

summarize-work