Agent skill

edge-verifier

Verifies relationships between a source text (e.g. Episode, Task) and existing Knowledge Graph lessons. Runs KNN/Hybrid search to find candidates, then uses LLM (scillm) to verify "verifies", "contradicts", or "related" stances with strict rationales.

Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/edge-verifier

Metadata

Additional technical details for this skill

short description
KNN + LLM verification for KG edges

SKILL.md

Edge Verifier

This skill links new content to the existing Knowledge Graph by:

  1. Recall: Running a KNN/Hybrid search (graph_memory.search) to find potential related lessons.
  2. Verify: Using scillm (LLM) to audit the relationship, classifying it as verifies, contradicts, or neutral.
  3. Link: Creating verified lesson_edges in ArangoDB.

Usage

Manual Execution

bash
# Verify edges for a specific source text
.pi/skills/edge-verifier/run.sh --source_id <ID> --text "Content to verify..."

# With explicit options
.pi/skills/edge-verifier/run.sh --source_id <ID> --text "..." --type "episode_turn"

Scheduling (Scheduler Integration)

This skill is designed to be scheduled via the /scheduler skill for continuous verification of new memory artifacts.

Register with Scheduler

bash
.pi/skills/scheduler/run.sh register \
  --name "edge-verify-nightly" \
  --cron "0 2 * * *" \
  --command ".pi/skills/edge-verifier/run.sh --batch" \
  --description "Nightly edge verification"

Jobs and logs are stored under ~/.pi/scheduler/. See .pi/skills/scheduler/SKILL.md for full options.

Prerequisites

  • .env must expose ArangoDB credentials.
  • CHUTES_API_KEY for LLM calls.

Expand your agent's capabilities with these related and highly-rated skills.

Didn't find tool you were looking for?

Be as detailed as possible for better results