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 agentmods add skills/joaoequer/oficina/client-facing-docsnpx skills add JoaoEquer/Oficina --skill client-facing-docsgit clone --depth 1 https://github.com/JoaoEquer/OficinaWhat 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 | $0.00065 | $0.00603 |
| Opus 5 | $0.00032 | $0.00302 |
| Sonnet 5 | $0.00013 | $0.00121 |
| Haiku 4.5 | $0.00006 | $0.00060 |
Grade A, and why
client-facing-docs 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 2d 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 — 42 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Client-facing documentation
A document for a client is not a technical document with a nice cover. It is a different kind of writing.
Principles
- Zero untranslated jargon. "Monorepo" becomes "a single repository". "Multi-tenant" becomes "each company with its data isolated". "CI/CD" becomes "every change passes automated checks before going live". If a technical term must appear, a one-sentence explanation comes with it.
- Honest numbers. Hours, deadlines and scope reflect the real contract — never a partial version that "presents better". A wrong number in presented material is the kind of thing that erodes trust and is very expensive to fix later.
- Document scope = contracted scope. Future phases, product vision and commercial possibilities stay out of the MVP document (at most, a one-line note: "future evolutions depend on a new commercial cycle").
- Real team, real roles. Only people actually allocated, each with their correct role. Double-check before sending.
- Explicit engineering principles. A short section stating how the team decides: proven technology over trends, complexity only when the problem demands it, reversible decisions, honest deadlines. Clients understand and value it — and the document becomes an anchor when someone suggests the trend of the week.
House formats
Architecture/structure document (responsive HTML):
- Sections: overview → what the system does → how it is organized (repos, modules) → team and roles → engineering principles → timeline/workstreams
- Flowing prose, tables only where comparison demands them
- No decorative Unicode borders and no emoji overload (they break rendering and age badly)
Presentation (fullscreen HTML deck):
- Keyboard (←/→) and click navigation, progress bar, slide counter
- One topic per slide, short sentences
- Export to PDF for distribution via WhatsApp/email (the PDF is what circulates; the HTML is what presents)
- Mandatory final check: deck numbers × contract numbers, before presenting
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.
- 2d ago First seen · 42 lines · 65 tokens per session scan A eb8b7368f985
client-facing-docs is a skill published in the GitHub repository JoaoEquer/Oficina (2 stars, last pushed 1mo ago), licensed MIT. It adds 65 tokens to every session and 603 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 2026-08-31.
Other skills, from other repositories
openrig-user
Use when a specific rig command, subcommand, or flag is already known and you need its exact syntax, JSON shape, defaults, or error meaning. NOT for natural capability discovery, open-ended how-do-I questions, or choosing which OpenRig move applies.
orienting-to-an-inherited-seat
Use when you have just been primed into an EXISTING seat through a planned handover — a different agent retired and handed you the seat plus its earned context — and you need a world model of what just happened to you. Covers how a handover differs from compaction and from a fresh launch, the…
forming-an-openrig-mental-model
Use when the system around you does not make sense yet: you just booted into a seat and do not know how the pieces fit; someone said rig, pod, seat, fleet, topology, or slice and you are not certain what they mean here; you are unsure what kind of rig you are in or what it is for; you do not know how skills reach you…
openrig-upgrade
Use when upgrading the OpenRig CLI/daemon on a host with running rigs, especially when preserving tmux-backed agent sessions through a daemon restart or documenting hot-upgrade SOP evidence.
openrig-skills
Use when you're operating OpenRig and need the right skill or context for fleet recovery, seat handover, new-seat orientation, a watchdog wake, cross-host reach to an agent on another machine, rig packaging, an OpenRig upgrade, systematic debugging, queue triage, or implementation planning; also use when you don't…
claude-compaction-restore
Use when a Claude Code session has just compacted, is about to compact, reached context limit, resumed after /compact, or needs to rebuild its working mental model from Claude JSONL transcripts and touched files.