Security audit
Security checks for vulnerabilities and agentic risk
This skill is a reference helper for Forrest Landry's Immanent Metaphysics and does not show hidden access, persistence, credential use, or destructive behavior.
Installers should understand that the skill may fetch public mflb.com pages when used and that some packaged reference files described by SKILL.md appear to be missing. The publisher should narrow trigger wording and replace absolute example paths with relative placeholders, but the inspected behavior is purpose-aligned and low risk.
The trigger list is broad and includes generic phrases like modality questions, ethics-related terms, and any request to ground claims in the framework. That can cause the skill to activate outside narrowly intended contexts, increasing the chance it overrides better-scoped skills or injects domain-specific guidance into unrelated conversations.
The example command exposes a concrete local filesystem path including a username and skills directory layout. While presented as documentation, this leaks environment structure that could aid skill enumeration, targeted probing, or adaptation of later attacks to a known host layout.
# Set path relative to skill dir (resolve from dirname of SKILL.md)
GRAPH="$(dirname $(realpath ~/Tillerman/Eitan/skills/im-framework/SKILL.md))/references/graph.jsonl"
# Search by concept name
grep -i '"name": "symmetry"' "$GRAPH"
This YAML file is a manifest-like reference file, so vague-trigger checks apply. The heading only says it is an ontology schema for extracting structured knowledge from Forrest Landry's work, but it does not define any explicit invocation boundaries, trigger phrases, or exclusion conditions, which could leave activation criteria underspecified in systems that rely on manifest/reference descriptions.
No suspicious patterns detected.