The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on the execution of local command-line binaries to manage tasks and report progress.
It executes bin/larc for task ingress, context retrieval, and state updates (e.g., bin/larc ingress openclaw, bin/larc ingress done).
It executes addness-cli to synchronize coding tasks with goal management (e.g., addness-cli work start, addness-cli progress).
[DATA_EXFILTRATION]: The skill transmits operational data to external platforms.
It sends task statuses and notes to the Lark/Feishu environment via the bin/larc utility and the openclaw-lark plugin.
It updates progress and PR URLs in the Addness goal tracking system.
[PROMPT_INJECTION]: The skill has a surface for indirect prompt injection due to the processing of external data.
Ingestion points: The commands bin/larc ingress context and bin/larc ingress handoff ingest data directly from the external LARC/Lark task queue into the agent's context (SKILL.md).
Boundary markers: There are no explicit delimiters or instructions provided to the agent to treat the ingested task data as untrusted or to ignore embedded instructions.
Capability inventory: The agent has access to shell execution via bin/larc and addness-cli, and network interaction through the openclaw-lark plugin.
Sanitization: No sanitization or validation logic is specified for the data retrieved from the task queue.

larc-runtime