Diagnose

Backward-looking root-cause analysis for /align corrections. Read an /align pass file + the project's instruction surfaces (CLAUDE.md, skills/*/SKILL.md, references/, TASKS.md, optionally smart-memory) and trace each ❌ / 🔶 / 🔷 claim back to the stale, contradictory, or missing instruction that caused it. Output: a per-claim trace report with quoted-evidence citations, 0–100 confidence scoring, and threshold-80 filtering. Source-agnostic by design: any /align pass file (whatever producer emitted the original claims) can be /diagnose'd against any project's instruction surfaces. Trigger words:

1 starsProductivity & Notes Plugin marketplace

Install

Claude Code plugin
/plugin marketplace add ggrigo/align

The repo is a Claude Code plugin marketplace. Run /plugin afterwards to pick the plugin that includes this skill.

skills CLI (Claude Code, Cursor, Codex and more)
npx skills add ggrigo/align --skill diagnose
Manual install
git clone --depth 1 https://github.com/ggrigo/align.git /tmp/align
mkdir -p ~/.claude/skills
cp -r /tmp/align/skills/diagnose ~/.claude/skills/diagnose

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# /diagnose — root-cause analysis on faulty context

## Purpose

`/align` captures *what's wrong* in an LLM output (per-claim ratings + reality notes). `/retro` synthesizes *what to build next* across many sessions (longitudinal patterns + feature roadmap). `/diagnose` answers a third question: *why did this specific session go wrong, and where is the fix?*

The trio:

| Direction | Command | Time-scale | What it answers |
|---|---|---|---|
| Capture | `/align` | now | What's in this output? |
| Backward | `/diagnose` (this skill) | single-session | Why did this go wrong? |
| Forward | `/retro` | longitudinal | What features take us forward? |
| Ship | `/retro apply` | longitudinal | Open PRs from accepted patches |

`/diagnose` reads one pass file at a time. It does not cluster across sessions (that's `/retro`'s job).

## When to invoke

The user explicitly invokes `/diagnose <pass-file>` after running `/align` on a session and noticing the corrections suggest a recurring or structural problem — i.e., the operator's gut says *"this kind of mistake keeps happening,"* and they want to know which instruction surface to fix.

Auto-invocation is not supported. `/diagnose` is opt-in because it reads multiple project files and consumes meaningful tokens.

## Argument parsing

```
/diagnose <pass-file> [--scope=<globs>] [--include-smart-memory] [--threshold=<N>]
```
Read the full skill on GitHub

Details

Skill name
diagnose
Repository
ggrigo/align
Path
skills/diagnose/SKILL.md
License
MIT
Repo updated
18 July 2026

FAQ

What does the Diagnose skill do?

Backward-looking root-cause analysis for /align corrections. Read an /align pass file + the project's instruction surfaces (CLAUDE.md, skills/*/SKILL.md, references/, TASKS.md, optionally smart-memory) and trace each ❌ / 🔶 / 🔷 claim back to the stale, contradictory, or missing instruction that caused it. Output: a per-claim trace report with quoted-evidence citations, 0–100 confidence scoring, and threshold-80 filtering. Source-agnostic by design: any /align pass file (whatever producer emitted the original claims) can be /diagnose'd against any project's instruction surfaces. Trigger words:

How do I install the Diagnose skill?

Claude Code plugin: /plugin marketplace add ggrigo/align. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Diagnose popular?

It lives in ggrigo/align, which has 1 GitHub stars and 0 forks, and is included in 1 curated skill list.

Related skills

Align

1

ggrigo/align

Epistemic feedback loop. Generate an interactive HTML form where the user rates LLM-produced claims as correct, wrong, almost, needs-nuance, can't-verify, or skipped — plus add notes and flag missing items. Submit downloads an align.md that Claude reads back to apply corrections to TASKS.md, smart-memory (guarded), and the per-session archive. Source-agnostic by design: any producer (rhythm, digest, briefing, todo, code review, financial recon, anything) that emits claims matching the claim shape can be aligned. Trigger words: "/align", "align", "give feedback", "correct the rhythm", "that's n

Productivity & NotesIn 1 list

Retro

1

ggrigo/align

Synthesize the /align archive into a pass document — aggregate metrics, failure-mode clusters, pipeline-blind-spot clusters, proposed patches with human-review gates. Reads the archive; never modifies it.

OtherIn 1 list

Cold Email Agency

1

CalebAnderson-Coder/ai-agency-os-starter

Write B2B cold emails and 4-step follow-up sequences that sell AI/automation services to US SMBs. Use when the user asks for cold outreach, a prospecting sequence, or reply-rate fixes.

Productivity & NotesIn 1 list

Audit Friction

5

TalonT-Org/AutoSkillit

Scan Claude Code project logs for friction patterns — repeated failures, approach loops, tool errors, misunderstanding cycles, and stuck workflows. Categorizes and counts friction events to surface what causes the most resistance. Use when user says "audit friction", "find friction", "friction audit", or "what keeps going wrong".

Productivity & Notes

Receipt Organizer

5

Dreydrey9000/claude-skills

Scan Gmail for receipts/invoices/payments from any month, create a labeled folder, and tag all receipt emails into it. Run monthly before sending to accountant.

Productivity & Notes

Diagnose CI

5

TalonT-Org/AutoSkillit

Diagnostic executor for CI failures. ALWAYS invoke this skill when instructed to diagnose CI failures. Do not fetch CI logs directly — use this skill first to load the diagnosis workflow.

Productivity & Notes

Browse the most popular Claude skills, all productivity & notes skills, or write your own with the skills generator.