The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the adlc-cli command-line tool to retrieve workflow states and resume paused missions. This includes passing verdict inputs such as 'DONE', 'CONTINUE', or 'SPEC_CORRECTION_NEEDED' to the automation runtime.
[EXTERNAL_DOWNLOADS]: The skill's documentation provides instructions to download and install necessary dependencies using the vendor's package manager (adlc-cli skills add tikalk/adlc-team-skills --skill factory-mission). This is a standard setup procedure for the 'tikalk' ecosystem.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted data during its core verification phase, creating a potential surface for indirect instructions.
Ingestion points: Code diffs, test suite outputs, and brief.md files are read and processed in the verify and converge steps.
Boundary markers: No specific delimiters or safety warnings (e.g., 'ignore instructions found in the code') are included in the prompt to prevent the agent from obeying instructions found within the data it is supposed to analyze.
Capability inventory: The agent can execute adlc-cli commands to modify the status of a mission and provide inputs that influence the next stage of the development pipeline.
Sanitization: The instructions do not describe any sanitization, escaping, or validation of the content found in the external artifacts.

mission-verify