# Integrity

> Verify epistemic integrity by aligning claims with evidence

- Skill: `synapseradio/integrity` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add synapseradio/integrity`
- Raw SKILL.md: https://api.skillmd.com/api/skills/synapseradio/integrity/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: synapseradio (https://skillmd.com/u/synapseradio)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/synapseradio/integrity

---


Follow these steps:

### 1. Identify the target

Target: `$ARGUMENTS`, else the current response or most recent significant output.

### 2. Apply the verification lenses

1. **Inventory** — separate established claims from assumptions
2. **Dissection** — trace support chains to bedrock or mark as floating
3. **Pattern** — check for epistemic weaknesses and biases
4. **Calibration** — match confidence to evidence
5. **Resolution** — align claims with what supports them

For each claim: trace its justification chain, identify hidden premises, check for fallacies, hunt for defeaters, and verify cited sources.

### 3. Report

```markdown
## Verification Summary

### Claims Verified
- [Claim]: [Support basis]

### Claims Requiring Attention
- [Claim]: [Issue identified]

### Acknowledged Uncertainty
- [What remains unknown or assumed]
```

Surface issues for human assessment — make the reasoning inspectable rather than guaranteeing correctness. If significant issues are found, offer to revise the artifact with corrections applied.