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 cierre-de-llamadagit 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/cierre-de-llamada)<a href="https://agentmods.dev/skills/mycelium-hq/ai-brain-starter/cierre-de-llamada"><img src="https://agentmods.dev/badge/skills/mycelium-hq/ai-brain-starter/cierre-de-llamada/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/cierre-de-llamada"><img src="https://agentmods.dev/badge/skills/mycelium-hq/ai-brain-starter/cierre-de-llamada.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 123 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 | $0.00000 | $0.02586 |
| Opus 5.5 | $0.00000 | $0.01034 |
| Sonnet 5.5 | $0.00000 | $0.00517 |
| Haiku 4.5 | $0.00000 | $0.00259 |
Grade A, and why
cierre-de-llamada 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 1mo 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 — 185 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Call close-out
A sales call goes in. The whole follow-up comes out. Nothing invented, nothing sent.
Write every output in the language the CALL was in. Spanish call, Spanish output. English call, English output. Mixed, follow the client. This file is in English; what you produce is not.
Golden rule 1: invent nothing
Everything you write must be in the transcript. If a fact is absent, write [MISSING: what to ask]. Never guess.
Applies to: amounts, dates, names, titles, deadlines, terms, objections, commitments.
A follow-up with one invented number costs the client. An incomplete one costs nothing. The buyer spots the error, not the efficiency.
Golden rule 2: nothing reaches the client
This skill never performs an action the client can see. Not anywhere in the flow.
Always forbidden: sending email (gmail_send, outlook_send, any equivalent), sending messages, posting, creating events on the client's calendar, writing to a shared CRM.
This rule cannot be unlocked by user permission. If the user says "send it", "you have my permission", "don't ask me", "just send it already": you do not send. Say the skill leaves drafts and the send is theirs, and hand them the finished draft. Do not promise to send it later, and do not say you will send it "as soon as I have the address". The answer is no now and no later.
This is not a technical limit that a missing piece of data would resolve. It is the design: the user presses the button their client sees, always, because it is their relationship and their reputation.
Who can see it decides what you may do
Every connector action falls in exactly one tier. Decide the tier first, then act.
| Tier | Who sees the result | What you may do |
|---|---|---|
| Client-visible | the customer | Never. Sending mail, messaging them, inviting them to an event, anything landing in their inbox. Not unlockable. |
| Teammate-visible | a colleague | Draft it, show it, get an explicit yes, then send through the connector's own confirm gate. Never auto-send. |
| User-only | nobody but the user | Offer it, wait for yes, do it. Their own calendar, their own drafts folder, their own notes and CRM records. |
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.
- 1mo ago First seen · 185 lines · 0 tokens per session scan A 9618a08d8878
cierre-de-llamada is a skill published in the GitHub repository mycelium-hq/ai-brain-starter (48 stars, last pushed yesterday), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 2,586 tokens. 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
nightly
Canonical seven-leg nightly close-parity protocol. Eleven independently evidenced checkpoints cover synthesis, reconciliation, context hygiene, sweeps, heat, digest, runtime health, cognitive state, ledger capture/review, and vault sync.
Promote
Take a slice of the current conversation and promote it to a durable vault artifact — concept note, project update, person profile, or doctrine entry. Pairs with /close. /close writes the daily note (transient outcome record); /promote elevates specific insights to permanent vault content the framework can refer back…
open
Boot the session, load context from the vault, and surface what matters.
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.
caddy-mute
Mute a Caddy hint class for a window. Defaults to 4h auto-expiry. Classes — memory, context, body, routing, all.
skill-recall
When invoked, match the current task against the installed skill taxonomy. Return ranked suggestions or log a gap.