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.
npx skills add monkilabs/opencastle --skill self-improvementgit clone --depth 1 https://github.com/monkilabs/opencastleWrote 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.
[](https://agentmods.dev/skills/monkilabs/opencastle/self-improvement)<a href="https://agentmods.dev/skills/monkilabs/opencastle/self-improvement"><img src="https://agentmods.dev/badge/skills/monkilabs/opencastle/self-improvement/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.
<a href="https://agentmods.dev/skills/monkilabs/opencastle/self-improvement"><img src="https://agentmods.dev/badge/skills/monkilabs/opencastle/self-improvement.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00070 | $0.00545 |
| Opus 5 | $0.00035 | $0.00272 |
| Sonnet 5 | $0.00014 | $0.00109 |
| Haiku 4.5 | $0.00007 | $0.00055 |
Grade A, and why
self-improvement 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 5d 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.
What it actually says
Self-Improvement Protocol
Core Rule
Retry with different approach and it works → document lesson immediately. File: .opencastle/LESSONS-LEARNED.md
Writing a Lesson
⛔ HARD GATE — Use the CLI. Do NOT edit LESSONS-LEARNED.md directly.
opencastle lesson --title "Short descriptive title" --category general --severity high \
--problem "What was observed" --wrong "Failing approach" --correct "Working solution" \
--why "Root cause"
Required: --title, --category, --severity, --problem · Optional: --wrong, --correct, --why
After writing: if lesson reveals gap in skill/instruction file, update that file too (prevents pitfall at source).
Workflow
- Search LESSONS-LEARNED.md for matching entries or similar errors.
- Attempt task with conservative flags/options informed by lessons.
- On failure: retry with modified approach (up to threshold); capture error details, context.
- On success: run
opencastle lessonto record working approach. - Verify:
tail -1 .opencastle/LESSONS-LEARNED.md— confirm entry has title, category, severity. If malformed → re-run with corrected flags. - If lesson indicates needed skill/instruction update: draft change; propose a PR.
Quick search example:
rg "missing CRON_SECRET" .opencastle/LESSONS-LEARNED.md || true
Categories & Severity
Valid --category and --severity values, tagging format: LESSON-CATEGORIES.md.
Quality Rules
- Include exact error messages, commands, tool parameters
- Always fill both
--wrongand--correct— the contrast is what makes the lesson usable - One lesson per entry; code blocks mandatory for commands
Anti-Patterns
Never duplicate an existing lesson · Never defer logging to end of session
Agent Memory
For expertise tracking, cross-session knowledge graphs, load agent-memory skill.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 5d ago First seen · 58 lines · 70 tokens per session scan A 27fa4a242665
self-improvement is a skill published in the GitHub repository monkilabs/opencastle (61 stars, last pushed 11d ago), licensed MIT. It adds 70 tokens to every session and 545 once invoked, about $0.0003 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 2026-09-03.
Other skills, from other repositories
product-architect
Complete product development system with 80 agents and 36 frameworks. Use when the user wants to build a product, write a PRD, plan an MVP or roadmap, design an app, research a market or check whether a feature already exists or is novel, do competitive analysis, run a security audit, build a financial model, plan…
workspace-brief
Build a concise, evidence-backed briefing from workspace status and recent commits.
agentic-delegation
Delegate exploration sweeps to Haiku subagents and bulk writing to Sonnet subagents while the orchestrator keeps architecture, security-sensitive edits, and commits; use at the start of any multi-step task in this repo to minimize token spend by delegating to cheaper models.
skillify
Auto-create new skills from repetitive workflows. When you notice yourself doing the same multi-step process repeatedly, extract it into a reusable SKILL.md that any agent can use.
aws-non-destructive-task-automation-advisor
Design AWS non-destructive task automation using EventBridge, Step Functions, Lambda, Systems Manager Automation, SNS, SQS, approvals, notifications, reporting, and evidence gathering. Use only for read-only or coordination-safe automation; do not use for destructive remediation or mutation-heavy runbooks.
handoff
Compacts current session state into a clean handoff document for a fresh agent session.