The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a Bash script in SKILL.md for version verification. It uses standard utilities like curl, grep, and awk to check for updates from the official repository. The script only echoes update instructions for the user and does not execute any remote payloads automatically.
[EXTERNAL_DOWNLOADS]: The skill fetches market data from api.barker.money and version metadata from raw.githubusercontent.com. These network operations are limited to the skill's stated purpose and use domains associated with the author's infrastructure.
[INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection by processing external data from a market API and a historical database.
Ingestion points: Stablecoin stress signals and TVL data are fetched dynamically from https://api.barker.money/api/public/v1/market/overview and included in responses.
Boundary markers: The skill contains a dedicated 'Security: External Data Boundary' section that explicitly instructs the agent to treat API values as data only, avoid following embedded commands, and render outputs verbatim.
Capability inventory: The skill performs network calls to specific vendor endpoints and maintains a local version cache in the user's home directory.
Sanitization: Risk is mitigated through instructional guardrails that warn the agent against executing imperative text found in external API fields.

stablecoin-depeg-monitor