Use when a persona is asked to self-improve, promote a repeated correction into a rule, turn recurring work into a skill, or review its local procedures.
Scanned 8/30/2026
Install to Claude Code
npx -y skills add kickinrad/personas-framework --skill self-improve --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Self Improve?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/kickinrad-self-improve)More formats (shields.io, HTML) on the badges page.
---
name: self-improve
description: Use when a persona is asked to self-improve, promote a repeated correction into a rule, turn recurring work into a skill, or review its local procedures.
---
# Self-improve
Improve the persona without duplicating authority.
## Route the change
- Identity, voice, role, and boundaries belong in `AGENTS.md`.
- A recurring multi-step procedure belongs in one portable role skill.
- Explicit persona memory belongs in ignored `user/memory/MEMORY.md` with user
approval.
- Stable knowledge belongs in its owning external system, with only the minimum
useful persona-local pointer.
- Claude- or Codex-only mechanics belong in the matching runtime adapter.
- Framework structure belongs in Personas and routes to `persona-dev`.
## Promote evidence
After repeated real friction, cite the persona-local evidence and propose one
concise rule, one reusable skill, an existing integration, or removal of stale
material. Promote the smallest rule that changes the observed behavior; do not
generalize one correction into fleet-wide policy. Wait for approval before
changing identity, rules, tools, skills, connections, or external services.
## Audit
Review the portable definition, explicit memory, profile, skills, adapters,
integrations, and loose files. Report proposed additions, updates, and removals
with evidence. Apply approved changes to one owning surface, inspect the exact
diff, and propose a persona-local commit.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!
Ultra-compressed communication mode. Cuts token usage ~75% by speaking like caveman while keeping full technical accuracy. Supports intensity levels: lite, full (default), ultra, wenyan-lite, wenyan-full, wenyan-ultra. Use when user says "caveman mode", "talk like caveman", "use caveman", "less tokens", "be brief", or invokes /caveman. Also auto-triggers when token efficiency is requested.
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', ...
**Complete production-ready guide for Google Gemini embeddings API** This skill provides comprehensive coverage of the `gemini-embedding-001` model for generating text embeddings, including SDK usage, REST API patterns, batch processing, RAG integration with Cloudflare Vectorize, and advanced use cases like semantic search and document clustering. ---
Interview, source-challenge, verify, save, and ADR-gate fuzzy coding requests into Codex-ready implementation specs. Use when a feature, bugfix, refactor, migration, repo-wide change, or architecture task needs user-verified requirements, source-backed decisions, durable architecture decisions, acceptance criteria, validation commands, rollout notes, saved spec/ADR files, and a Codex execution prompt. Do not use when already fully specified or when the user wants direct implementation now.
Use when a repo needs CodeGraph plus ast-grep for Codex MCP setup, exploration, impact analysis, structural search, or safe refactor planning.