From grimoire
Deduplicates security findings by indexing, comparing pairs for duplicates or similarities, presenting tables with recommendations, and executing user-confirmed deletions or merges.
npx claudepluginhub joranhonig/grimoireThis skill uses the workspace's default tool permissions.
Identify and resolve duplicate or overlapping security findings.
Reviews existing security findings against best practices for title clarity, description completeness, recommendation objectivity, severity accuracy, and reference validity. Validates structure with bash script.
Coordinates parallel code reviews across dimensions like security, performance, architecture, testing, and accessibility with finding deduplication, severity calibration, and consolidated reports.
Consolidates outputs from dev, design, writing, product, and UX detect skills into unified scored index, risk heatmap, and unknowns backlog for audit combination.
Share bugs, ideas, or general feedback.
Identify and resolve duplicate or overlapping security findings.
Before starting, read skills/finding/SKILL.md to understand finding structure and
conventions. That skill defines the format and standards that this workflow relies on for
comparison.
When this skill is activated, create a todo list from the following steps. Mark each task in_progress before starting it and completed when done.
- [ ] 0. Load finding knowledge (read skills/finding/SKILL.md)
- [ ] 1. Index and compare findings
- [ ] 2. Present duplicates and confirm actions
- [ ] 3. Execute and report
Read skills/finding/SKILL.md to internalize finding structure and conventions. This is
required before proceeding — the base skill defines the format you will compare against.
Run the indexing script to get the full finding set:
bash skills/finding/scripts/index-findings.sh
If fewer than 2 findings exist, report that there is nothing to deduplicate and stop.
For each pair of findings, compare title, type, context (affected files), and description. Classify each pair as:
If the finding set is large (>10 findings), group by type first and only compare within
groups. Use subagents for parallel comparison if needed.
See skills/finding/examples/dedup-scenario.md for a worked example showing duplicate vs
similar classification and the resolution workflow.
Present results:
For each duplicate pair: "Delete <file>? [y/n]"
For each similar pair: "Merge into <file>? [y/n/skip]"
Never delete or merge without explicit user confirmation.
Perform confirmed deletions and merges. When merging:
Re-run the index to show the updated finding set:
bash skills/finding/scripts/index-findings.sh
Suggest /finding-review on any merged findings to verify quality.