The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from repositories, git history, and system logs, creating an attack surface where embedded malicious instructions in those files could influence the agent's behavior.
Ingestion points: Extensive use of ripgrep (rg), find, and git grep to scan the local filesystem and project history. (File: references/detailed-guide.md)
Boundary markers: The skill instructions specify that the agent should never print raw secrets and should use redacted fingerprints, but specific prompt delimiters for scan results are not defined. (File: SKILL.md)
Capability inventory: The agent has access to powerful tools including ripgrep, git, docker, and python3. (File: references/detailed-guide.md)
Sanitization: The guide emphasizes redaction in logs but lacks clear sanitization or escaping protocols for interpolating external file content into the agent's prompt context.
[COMMAND_EXECUTION]: The skill defines a wide array of shell commands to inspect the local environment and manage system settings.
Ripgrep patterns are used to search for high-entropy strings, private keys, and specific provider tokens (OpenAI, AWS, Stripe). (File: references/detailed-guide.md)
Use of git rev-list and git grep to perform deep inspection of all commits in the repository history. (File: references/detailed-guide.md)
The guide describes firewall management using ufw or iptables and service management via systemd. (File: references/detailed-guide.md)
[DATA_EXFILTRATION]: The skill is explicitly designed to find and access sensitive credential files and environment metadata.
A Python one-liner (python3 -c 'import os,re; ...') is used to extract and list the names of environment variables matching sensitive patterns like 'openai', 'token', or 'password'. (File: references/detailed-guide.md)
The skill targets sensitive paths such as ~/.ssh/id_rsa, .env files, and JSON service account keys. (File: references/detailed-guide.md)
[PERSISTENCE]: The guide includes instructions for establishing persistent components on the host system.
Mentions the creation and management of systemd services for API gateways and applications in /etc/systemd/system/. (File: references/detailed-guide.md)
[DYNAMIC_EXECUTION]: The skill utilizes dynamic script execution for environment analysis.
Execution of Python code at runtime to filter and report on environment variables. (File: references/detailed-guide.md)

cred-omega