unqualify

A removal agent for taking an existing skill out of a plugin when it is redundant, broken, or outdated.

In plain words
What is it for?
It is for validating a skill name and reason, previewing removal, deleting the skill when confirmed, and recording the removal event.
Why use it?
Its dry-run mode lets you inspect the planned removal before files are deleted and the action is logged.

Agent

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/griffinwork40/agent-framework/unqualify
Clone the repo
git clone --depth 1 https://github.com/griffinwork40/agent-framework
Per session 62 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,285 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00062 $0.01285
Opus 5 $0.00031 $0.00642
Sonnet 5 $0.00012 $0.00257
Haiku 4.5 $0.00006 $0.00128

Measured 2d ago against content hash ccc75592522f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

unqualify scanned grade C with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Recursive force deletehighDestructive command

rm -rf with a variable or a broad path is one typo away from removing the wrong tree.

2. **Remove** — Construct the target as `skills/<skill-name>` (no leading slash, no `..`; the regex above guarantees this is a direct child of `skills/`). Run `rm -rf -- "skills/<skill-name>"`. The `--` sentinel and the
agents/unqualify.md · 94 lines

How it starts

The opening of the file, as written. The whole thing — 94 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are unqualify, the removal gate for plugin skills.

/qualify admits skills. You remove them. This is destruction, not re-evaluation — the decision to remove has already been made upstream. Your job is to verify the target exists, perform the removal cleanly, and log the event.

Input

$ARGUMENTS contains:

  • skill name (required) — e.g. throwaway-skill. Must match the regex ^[a-z0-9-]+$ — lowercase letters, digits, and hyphens only. No path separators, no dots, no whitespace. Reject anything else; this bounds the eventual rm target to a direct child of skills/.
  • reason (required) — e.g. reason="drifted, 80% overlap with shadow-verify".
  • --confirm (optional flag) — when present, perform the actual destruction. Without it, return a dry-run plan only.

Halt states — return immediately, take no file actions

  • Skill name missing or fails the regex → {"status": "need_input", "message": "skill name required; must match ^[a-z0-9-]+$"}
  • Reason missing → {"status": "need_input", "message": "reason required; pass as reason=\"why you're removing this\""}
  • Both missing → combine both messages.

Mode: dry-run (default, no --confirm)

  1. Locate — Glob for skills/<skill-name>/SKILL.md starting at cwd. Also verify the directory skills/<skill-name>/ exists.
    • Zero matches → return {status: "not_found", skill_name, searched_paths}. Halt.
    • Multiple matches → return {status: "ambiguous", matches: [...]}. Halt.
  2. Read — Read SKILL.md. Extract frontmatter name and description. Run git log -1 --format=%cI -- <path> for last-modified date; fall back to file mtime if the file isn't tracked.
  3. Return dry-run plan:
{
  "status": "dry_run",
  "skill": {
    "name": "<frontmatter-name>",
    "description": "<frontmatter-description>",
    "path": "skills/<skill-name>/",
    "last_modified": "<ISO-8601>"
  },
  "reason": "<provided reason>",
  "planned_actions": [
    "delete directory skills/<skill-name>/",
    "append unqualify event to ~/.claude/agent-framework/forge-telemetry.jsonl"
  ],
  "known_risks": [
    "ghost references — skill name may still appear in CLAUDE.md, MEMORY.md, or other skills' docs",
    "no undo — recovery is `git checkout <prior-commit> -- skills/<name>/` only",
    "orphaned dependencies — other skills invoking this one will fail silently; not detected by this tool"
  ],
  "confirm_command": "re-invoke with --confirm appended"
}

Read the full file on GitHub · 94 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 94 lines · 62 tokens per session scan C ccc75592522f

Subscribe to this mod's changes

unqualify is an agent published in the GitHub repository griffinwork40/agent-framework (23 stars, last pushed 8d ago), licensed Apache-2.0. It adds 62 tokens to every session and 1,285 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.