The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides structured instructions for the agent to use the ae-cli kb command-line tool to perform read-only operations like listing, indexing, searching (grep), and reading document pages. These commands are used to provide business context for data analysis.
[PROMPT_INJECTION]: The skill includes instructions to prevent the agent from following embedded instructions in external data by emphasizing that conclusions must be backed by evidence and that knowledge base retrieval is a supporting workflow, not the primary task.
Ingestion points: Content is ingested from internal knowledge bases via the +read and +grep subcommands of ae-cli kb (SKILL.md).
Boundary markers: The instructions provide explicit guidance to treat internal facts as optional context and to distinguish between analytical facts and retrieved evidence (SKILL.md).
Capability inventory: The agent has the capability to execute shell commands using the ae-cli tool (SKILL.md).
Sanitization: The agent is instructed to copy paths character-for-character and treat identifiers as opaque to prevent path manipulation (SKILL.md).
[DATA_EXFILTRATION]: While the skill accesses internal data, it contains a 'Safety Boundaries' section that explicitly prohibits the exposure of internal root paths, access tokens, or raw permission metadata. It also restricts the agent from performing destructive actions like deleting or modifying knowledge bases.

ae-kb-discovery