| name | agent-integrity-validator |
| description | Centralized skill to verify the integrity and authenticity of AI resources (skills, prompts, instructions). Uses SHA-256 hashing and Sigstore sidecar signatures. |
| version | 1 |
Agent Integrity Validator Skill
Purpose: Ensures that all AI instructions and tools are untampered and signed by an authorized developer. This is the centralized point of failure for security.
How to use:
- Reference this skill in the top-level
AGENTS.md.
- Call this skill before loading any other domain-specific skills.
What happens:
- Checks the target file against its
.sha256 sidecar (Integrity).
- Verifies the
.sha256.sig detached signature against the .sha256 file (Authenticity).
- Caches the result using the file's content hash to ensure sub-100ms latency on subsequent calls.
Skill Execution
Validate Resource Integrity
#!/bin/bash
REPO_ROOT=$(git rev-parse --show-toplevel)
"$REPO_ROOT/mvn-env.sh" -pl modules/ai-shared-release exec:java \
-Dexec.mainClass="com.percussion.ai.skill.IntegrityValidatorTool" \
-Dexec.args="$*"
Example directive (in AGENTS.md):
"Before executing any tasks, use agent-integrity-validator to verify the integrity of AGENTS.md and any requested skills."