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 mycelium-hq/ai-brain-starter --skill skillify-meta-loopgit clone --depth 1 https://github.com/mycelium-hq/ai-brain-starterWrote 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/mycelium-hq/ai-brain-starter/skillify-meta-loop)<a href="https://agentmods.dev/skills/mycelium-hq/ai-brain-starter/skillify-meta-loop"><img src="https://agentmods.dev/badge/skills/mycelium-hq/ai-brain-starter/skillify-meta-loop/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/mycelium-hq/ai-brain-starter/skillify-meta-loop"><img src="https://agentmods.dev/badge/skills/mycelium-hq/ai-brain-starter/skillify-meta-loop.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Excessive Agency · line 63 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
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 | $0.00111 | $0.01116 |
| Opus 5 | $0.00056 | $0.00558 |
| Sonnet 5 | $0.00022 | $0.00223 |
| Haiku 4.5 | $0.00011 | $0.00112 |
Grade A, and why
skillify-meta-loop 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 9d 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.
How it starts
The opening of the file, as written. The whole thing — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Bug → durable skill in 10 steps. The loop runs same-session, no queueing.
When to invoke
| Signal | Example |
|---|---|
| Same issue 2+ times | Two clients hit the same install error |
| Manual repeat | Wrote the same onboarding email 3× |
| Codified rule violated | A typo slipped past a lint rule despite it being codified |
| Support question pattern | 3 teams asked "how do we wire X" |
| Delivery friction | Bespoke setup hour you shouldn't be spending each time |
If the trigger is a one-off accident with no pattern signal, skip skillify. The loop is for repeats.
The 10 steps
1. Scaffold
Pick the skill's home:
- Private skills repo →
~/dev/<your>-skills/<name>/ - Public substrate →
<repo>/skills/<name>/(if generic + no personal data) - Team repo →
~/dev/<team>-skills/<name>/
mkdir <path> && touch <path>/SKILL.md. Don't symlink yet.
2. State the trigger
In SKILL.md frontmatter trigger: field: every phrase a user might type that should fire this skill. Cover bilingual surfaces if relevant. Cover the explicit invocation (/skill-name) AND natural-language patterns. Be over-inclusive; pruning false-positives is cheaper than missing the signal.
3. Define the failure mode
What does the world look like WITHOUT this skill? One paragraph. Be specific: name the file, the rule, the client, the hour cost. Without this, the skill ages into a "feels useful but no one runs it" file.
4. Write the steps in caveman form
Terse, IF/THEN, bullets, no preamble. Drop articles. Drop multi-paragraph rationale. The skill must be USABLE on read, not appreciable.
5. Provide one real example
Concrete, with real file paths, real flags, real outputs. Not <your-key-here>. Not # do the thing. The first real example is the most-cited part of the skill.
6. Add the discoverability path
Pick ONE: top-level SKILL.md (auto-loads metadata); plugin marketplace registration; hookify trigger reminder; CLI tool with explicit invocation. Document which path is wired. A skill nobody can find = a skill that doesn't exist.
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.
- 9d ago First seen · 69 lines · 111 tokens per session scan A 06b39e1548ff
skillify-meta-loop is a skill published in the GitHub repository mycelium-hq/ai-brain-starter (36 stars, last pushed 2d ago), licensed MIT. It adds 111 tokens to every session and 1,116 once invoked, about $0.0006 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.
Other skills, from other repositories
wiki-lint
Run a deterministic, read-only health check on an Obsidian wiki. Use for lint, vault health check, audit wiki health, find orphans, find dead links, frontmatter audit, provenance audit, or wiki audit. Reports graph, link, frontmatter, provenance-ledger, empty-section, and stale-index findings; it does not reason…
flow
Personal task and agent-session manager. CLI binary is flow (assumed on PATH) and stores metadata in /.flow/flow.db (SQLite). Use this skill when the user asks about their work, tasks, or projects in any natural phrasing — including but not limited to: "what's left", "what's remaining", "what's pending", "what do I…
log-triage
Parse bot logs and Langfuse traces into a short incident digest — what broke, how often, and the likely cause. Use when the user asks why the bot errored, wants a summary of recent failures, is debugging a crash, or after heartbeat-monitor reports a red check.
daily-brief
Compose a morning briefing for the user by pulling today's calendar, open tasks, and relevant memory, then delivering a short prioritized summary over the primary chat channel. Use when the user asks for a daily brief, "what's on today", a standup, or sets up a scheduled morning digest.
Stuck
Walk the 6-rung escalation ladder before declaring stuck or asking the principal. Most "stuck" is actually "haven't tried the next approach." Stopping is the last resort, not the first response.
tt
Quality gate immediately after building: prove that what was just built actually works, while the context is still hot. Six steps: scope what changed, run live on real data, break it on purpose, show a counter or log proving it ran, root-cause any failure and re-run, then a verdict with evidence. Only a green verdict…