The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using the ntm CLI tool to manage swarms, specifically using ntm spawn to initialize agents and ntm kill to terminate them.
[EXTERNAL_DOWNLOADS]: It invokes the escalate skill and deep-research-prompt to fetch external data and conduct research via models like GPT-5 Pro to stress-test conceptual premises.
[PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it processes untrusted data from the wiki and outputs from third-party agents.
Ingestion points: Reads content from _concepts/, _ops/focus-sweeps/, and agent-generated files such as WIZARD_IDEAS_*.md.
Boundary markers: Absent; the instructions do not implement explicit delimiters or "ignore" directives when reading concept files or agent ideation outputs.
Capability inventory: Includes management of external agent swarms (ntm), writing to the file system, and calling external research skills.
Sanitization: No filtering or validation is performed on the data ingested from the vault before it is used to drive the adversarial synthesis logic.

wiki-forge