Claude-Mind: Skill for Claude Code

.claude/skills/reflect-on-self/SKILL.md

reflect-on-self is a skill for Claude Code from zkysar1/Claude-Mind. It costs 96 tokens per session (4,872 once invoked), scanned A, original, MIT.

A skill for reviewing an AI agent's past outcomes to find reusable strategies and update its self-model. It also checks whether the agent's confidence matched how often its conclusions were correct.

In plain words
What is it for?
Use it during reflection cycles to extract patterns from resolved hypotheses, build strategic guidance, and calibrate confidence scores.
Why use it?
It turns individual results into broader lessons and can reveal when the agent is too confident or not confident enough.

Skill for Claude Code

Written for Claude Code: user-invocable in frontmatter.

This is zkysar1/Claude-Mind's own configuration. It tells Claude Code how to work on Claude-Mind itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything Claude-Mind configures →

Needs its repository: it runs a file that does not travel with it, so clone the repository first. The line is strategy_node = bash core/scripts/tree-find-node.sh --text "{strategy condition category or domain}" --leaf-only --top 1.

Reuse

Borrowing it

Nothing to install: this file belongs to zkysar1/Claude-Mind. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/zkysar1/Claude-Mind/main/.claude/skills/reflect-on-self/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/zkysar1/Claude-Mind

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for reflect-on-self

README.md
[![agentmods](https://agentmods.dev/badge/skills/zkysar1/claude-mind/reflect-on-self/github.svg)](https://agentmods.dev/skills/zkysar1/claude-mind/reflect-on-self)
Your own site
<a href="https://agentmods.dev/skills/zkysar1/claude-mind/reflect-on-self"><img src="https://agentmods.dev/badge/skills/zkysar1/claude-mind/reflect-on-self/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for reflect-on-self

Your own site · 80×15
<a href="https://agentmods.dev/skills/zkysar1/claude-mind/reflect-on-self"><img src="https://agentmods.dev/badge/skills/zkysar1/claude-mind/reflect-on-self.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 96 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 4,872 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
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 5.1 $0.00096 $0.04872
Opus 5 $0.00048 $0.02436
Sonnet 5 $0.00019 $0.00974
Haiku 4.5 $0.00010 $0.00487

Measured 12d ago against content hash add2b1491705, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

reflect-on-self scanned grade A with 0 findings 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 12d 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.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/skills/reflect-on-self/SKILL.md · 416 lines

How it starts

The opening of the file, as written. The whole thing — 416 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/reflect-on-self — Self-Model Reflection

This sub-skill implements self-model reflection modes for /reflect. It is invoked by the parent router for two modes:

  • Patterns mode (--extract-patterns): Mine resolved hypotheses for strategies, synthesize Level 1 patterns, build Level 2 strategic self-model
  • Calibration mode (--calibration-check): Analyze confidence calibration across all hypotheses

Each mode section below is self-contained with its own step numbering.


Mode: Extract Patterns (--extract-patterns)

This sub-skill implements Mode 2 of /reflect. It is invoked by the parent /reflect router when --extract-patterns is specified, or during --full-cycle after individual hypothesis reflections complete. It mines the full resolved population — --stage resolved AND --stage archived, per Step 1 — for reusable strategies, synthesizes Level 1 patterns and Level 2 strategic self-models, and updates the knowledge base.

Step 0: Load Conventions

Step 0: Load ConventionsBash: load-conventions.sh with each name from the conventions: front matter. Read only the paths returned (files not yet in context). If output is empty, all conventions already loaded — proceed to next step.

Step 1: Load the Full Resolved Population (resolved AND archived)

# BOTH stages are REQUIRED. `--stage resolved` alone is a SURVIVORSHIP FILTER:
# `resolved` is the small live holding area and records migrate to `archived`
# as they age, so most scoreable records sit in `archived`. Measured
# 2026-08-04T03:33 (bravo, hostname cc-05, uname -r 6.8.0-136-generic):
# resolved 86 vs archived 829 — resolved is 9.4% of the 915-record store.
# Pattern MINING is the worst possible consumer of that filter: a "recurring
# pattern" needs 3+ instances (Step 2), and mining 9.4% of the corpus both
# suppresses real patterns below threshold and lets a small recent cluster
# read as one. The `frequency: N` field emitted below, and the "Category
# accuracy clusters" in the pattern list, are computed straight off this
# fetch. (g-115-4866; template at review-hypotheses Mode 3 Step 1, g-115-3594.)
Bash: pipeline-read.sh --stage resolved
Bash: pipeline-read.sh --stage archived
# Mine the UNION of the two. Keep EXPIRED / UNRESOLVABLE out of any rate the
# patterns below assert (store-wide convention: only CONFIRMED and CORRECTED
# are scoreable) — but they remain legitimate inputs to non-rate pattern
# mining, so do not drop the records themselves.
Read all Level 0 reflections from journal entries
Read existing pattern files from $WORLD_DIR/knowledge/patterns/

Read the full file on GitHub · 416 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. 12d ago First seen · 416 lines · 96 tokens per session scan A add2b1491705

Subscribe to this mod's changes

reflect-on-self is a skill published in the GitHub repository zkysar1/Claude-Mind (5 stars, last pushed yesterday), licensed MIT. It adds 96 tokens to every session and 4,872 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.