Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsCommunityBlog
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Authors
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

Skill Proposer

ASecurity

Use when the user asks to turn wiki patterns into a concrete SKILL.md change, propose one atomic skill patch, or record Accept/Reject in the skill-impact ledger. Triggers on "propose a skill fix", "patch this skill from the wiki", "skill-proposer", or "gate a skill change".

10 stars
0 votes
0 copies
0 views
Added 9/28/2026
ai-agentspythonshellbash

Works with

cli

Security Analysis

A100/100

Scanned 9/28/2026

Install to Claude Code

$npx -y skills add fritzprix/libr-agent --skill skill-proposer --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Skill Proposer?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Skill Proposer
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/fritzprix-skill-proposer/badge)](https://www.skillsdirectory.com/skills/fritzprix-skill-proposer)

More formats (shields.io, HTML) on the badges page.

Files
SKILL.md
---
name: skill-proposer
description: >
  Use when the user asks to turn wiki patterns into a concrete SKILL.md change,
  propose one atomic skill patch, or record Accept/Reject in the skill-impact
  ledger. Triggers on "propose a skill fix", "patch this skill from the wiki",
  "skill-proposer", or "gate a skill change".
---

# Skill Proposer

Read the **host-global** skill-evolution wiki and propose **one** atomic skill
change. Works on Windows, macOS, and Linux. Meta workflow only — do not inject
wiki files into a normal task agent.

## Prerequisites

- Prefer the bundled **Wiki Maintainer** assistant (history + workspace for wiki CLI).
- On scheduled/autonomous wakes: **Propose** only — never Accept a skill patch
  without an explicit interactive user Accept (see wiki-maintainer
  `references/autonomous-loop.md`).
- Shell/Python to run **wiki-maintainer**’s CLI (shared wiki IO).
- Wiki root: `Path.home() / ".libragent" / "wiki"` (Linux/macOS `~/.libragent/wiki`, Windows `%USERPROFILE%\.libragent\wiki`).
- Resolve **wiki-maintainer** Base Directory as `<wiki-maintainer-dir>` (system skill install of `wiki-maintainer`).
- Editable target skill under a **runtime** skill root (user / assistant / workspace). Read [proposal-checklist.md](references/proposal-checklist.md).

```bash
python "<wiki-maintainer-dir>/scripts/wiki_cli.py" init
python "<wiki-maintainer-dir>/scripts/wiki_cli.py" path
```

## Not this skill

| Need | Use |
| --- | --- |
| Mine sessions → patterns | **wiki-maintainer** |
| Knowledge graph facts | **knowledge-distiller** |
| Brand-new skill with no wiki | Manual authoring / skill-creator guidance |

## Workflow

### 1. Read prior art

```bash
python "<wiki-maintainer-dir>/scripts/wiki_cli.py" cat index.md
python "<wiki-maintainer-dir>/scripts/wiki_cli.py" list-patterns
python "<wiki-maintainer-dir>/scripts/wiki_cli.py" cat patterns/<id>.md
python "<wiki-maintainer-dir>/scripts/wiki_cli.py" cat skill-impact.md
```

Skip proposals that already **Reject**ed the same change without new evidence. One pattern id per proposal.

### 2. Choose the owning skill

1. Map pattern direction to one skill name.
2. Edit a writable runtime skill root. If only a read-only system copy exists, propose overlay text for user/assistant skills — do not invent fake paths.

### 3. Draft one atomic change

1. Change one skill’s `SKILL.md` (or one reference file) only.
2. Keep frontmatter `name` + trigger-bearing `description`.
3. One-sentence **user-facing mechanism**.
4. Forbidden: evaluation answers, grader hints, checkout-only paths, machine-locked absolute paths in guidance (home-relative wiki paths via `wiki_cli path` are OK).

### 4. Validate

1. Check relative `references/` links inside the edited skill.
2. Self-check [proposal-checklist.md](references/proposal-checklist.md).
3. Prefer a short user-like smoke of the same failure class.

### 5. Ledger

```bash
python "<wiki-maintainer-dir>/scripts/wiki_cli.py" append-impact \
  --skill "<name>" \
  --summary "<what changed>" \
  --mechanism "<user-facing mechanism>" \
  --validation "<what you ran or Pending>" \
  --decision Pending|Accept|Reject \
  --notes "<optional>"
```

On **Reject**, restore the skill content. **Never** delete wiki pattern or ledger rows.

### 6. Hand-off

Summarize: pattern id, skill, mechanism, decision, remaining Pending work.

## Safety

- One proposal at a time; new evidence before retrying Reject.
- Wiki is global per OS user account; skills stay in skill roots.

Attribution

fritzprixfritzprix
View sourceMore from fritzprix →
SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Related Skills

Caveman

Ultra-compressed communication mode that cuts output tokens while keeping technical accuracy. Levels: lite, full, ultra and the wenyan variants. Use for /caveman, "caveman mode", "talk like caveman", "be brief" or "less tokens".

1074701 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

695601 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3351 votes

catchup

Recovers the conversation and failed tool calls of a previous Codex, Claude Code, Antigravity, Cline, Copilot CLI, Cursor, DeepSeek Harness, Kimi, OpenCode, Pi Agent, or ZCode session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", asks to recover/summarize a previous session before continuing, or asks to diagnose or report a catchup failure. Do NOT use for the current conversation, git history, or any non-agent log.

691 votes

math-skill

A comprehensive mathematical reasoning skill for AI assistants — handles arithmetic to research-level problems with rigorous step-by-step reasoning, systematic verification, and transparent uncertainty handling

381 votes
View all in ai-agents →