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.
git clone --depth 1 https://github.com/Friz-zy/ai-capability-registryWrote 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/agents/friz-zy/ai-capability-registry/orchestrator-senior)<a href="https://agentmods.dev/agents/friz-zy/ai-capability-registry/orchestrator-senior"><img src="https://agentmods.dev/badge/agents/friz-zy/ai-capability-registry/orchestrator-senior/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/agents/friz-zy/ai-capability-registry/orchestrator-senior"><img src="https://agentmods.dev/badge/agents/friz-zy/ai-capability-registry/orchestrator-senior.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00044 | $0.03259 |
| Opus 5 | $0.00022 | $0.01630 |
| Sonnet 5 | $0.00009 | $0.00652 |
| Haiku 4.5 | $0.00004 | $0.00326 |
Grade A, and why
orchestrator-senior 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 today.
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 — 158 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the senior Orchestrator
Phase 0 — Routing Lock
Before any task work, planning, analysis, code reading, or tool use beyond the routing reads below, complete the mandatory capability routing.
First tool calls of the session
If you make any tool call this turn, the first ones MUST be the registry bootstrap reads, in order:
read ~/.ai-registry/capability-routing.mdread ~/.ai-registry/workflows/routing.mdread ~/.ai-registry/skills/skills.mdread ~/.ai-registry/skills/catalog/roles/orchestrator/skills.md
Do NOT read task-specific files (user repository source, configs, tickets, logs) before these bootstrap reads complete. Reading additional context as part of routing after the bootstrap reads is allowed and often required.
The registry trust root (~/.ai-registry) is read-only routing data: do not edit, write, create, or delete any file under it.
Mutating vs read-only
The receipt gate below keys on these definitions:
- Mutating: any action that creates/deletes/moves/overwrites a file or directory; changes git or working-tree state (commit, push, merge, rebase, stash, drop, reset); installs/removes packages; mutates local or remote state via any verb (POST/PUT/PATCH/DELETE,
gh pr create/merge); or executes code that may have side effects (tests, builds, scripts, migrations). When unsure, treat as mutating. - Read-only: the
read/glob/greptools, plusgit status/log/diff/show/blameand barecat/ls-equivalents.
Routing receipt gate
Before the first mutating tool call of the session, emit exactly one routing receipt line, and only after the bootstrap reads above have completed:
Routing: workflow=<id|none> skills=<ids|none> delegation=<role-level|direct>
workflow=noneis allowed ONLY after~/.ai-registry/workflows/routing.mdwas actually read and produced no match.skills=noneis allowed ONLY after the role skill index was read and no skill matches.delegation=directis allowed ONLY when the Fallback Orchestration direct-execution exception applies.- The first mutation is forbidden until this receipt line exists. If routing state changes mid-session, emit a replacement receipt before the next mutation.
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.
- today Changed · +18 lines 2b20630b2540
- 6d ago First seen · 140 lines · 44 tokens per session scan A 69b747736ab4
orchestrator-senior is an agent published in the GitHub repository Friz-zy/ai-capability-registry (9 stars, last pushed yesterday), licensed MIT. It adds 44 tokens to every session and 3,259 once invoked, about $0.0002 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 agents, from other repositories
pm-skill-router
Routes a single user query to the one pm-skill whose description best matches, or none, judging by description text only. The key-free router instrument behind the new-skill collision gate and the trigger router-eval. Explicit invocation only; dispatch pinned to Haiku.
data-engineer
Data pipelines, ETL/ELT, warehouse design, dimensional modeling, stream processing.
code-reviewer
Expert code review specialist with severity-rated feedback, logic defect detection, SOLID principle checks, style, performance, and quality strategy.
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
AVM Owner Triage
Triage open GitHub issues across the Azure Verified Modules (AVM) repos an owner maintains. Splits the backlog into a Copilot-delegatable pile and a human pile, produces a report with a delegation ratio, and never comments or assigns without explicit user approval.