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/logicrw/agpair/hermesnpx skills add logicrw/agpair --skill hermesgit clone --depth 1 https://github.com/logicrw/agpairWrote 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/logicrw/agpair/hermes)<a href="https://agentmods.dev/skills/logicrw/agpair/hermes"><img src="https://agentmods.dev/badge/skills/logicrw/agpair/hermes.svg" alt="Measured on agentmods" 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 | $0.00027 | $0.02014 |
| Opus 5 | $0.00014 | $0.01007 |
| Sonnet 5 | $0.00005 | $0.00403 |
| Haiku 4.5 | $0.00003 | $0.00201 |
Grade A, and why
agpair 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 3d 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 — 274 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGPair for Hermes
Use AGPair as Hermes' durable external-agent control plane. Hermes remains the controller, verifier, and final decision maker.
Core rule: external-first, not external-always. For non-trivial coding, refactor, test-fix, research, review, or multi-file work, consider an AGPair lane before doing everything directly in Hermes. Skip AGPair when the task is tiny, highly sensitive, depends on hidden Hermes context, or requires local home-state inspection that should not be delegated.
Hermes is not an AGPair executor identity. Do not inherit Codex or Claude Code self-executor suppression. For Hermes controllers, all active external AGPair executors are valid:
grok-cliantigravity-clicodexclaude-code
Inspect the live policy before relying on exact ordering:
agpair policy list --controller hermes --json
If the policy and this skill disagree, the live AGPair policy is the execution
truth. Prefer explicit --executor when you want a specific lane.
When To Use
Use AGPair when an external worker can produce useful evidence:
- implementation candidates in isolated worktrees;
- independent code review or adversarial critique;
- repo scans, test-fix attempts, and repetitive verification;
- research or diagnosis with a bounded question;
- role-distinct fanout where each lane has a different scope, hypothesis, or review angle.
Use Hermes directly, or a Hermes native subagent, when the work is:
- a tiny local edit or one-command check;
- sensitive credential, auth, config, or memory inspection;
- dependent on current hidden conversation state;
- not safely isolatable for mutating external work;
- best handled as controller-side synthesis or final review.
AGPair-started workers do not share Hermes' hidden session state. Every AGPair brief must stand alone: include the goal, repo path, allowed scope, forbidden scope, constraints, required output, and validation/evidence requirements.
Routing Budget
Before dispatching, choose a routing budget:
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.
- 3d ago First seen · 274 lines · 27 tokens per session scan A b0d6ce9ea506
agpair is a skill published in the GitHub repository logicrw/agpair (48 stars, last pushed 2mo ago), licensed MIT. It adds 27 tokens to every session and 2,014 once invoked, about $0.0001 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 2026-08-30.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
auto-perf-optimize
Run agent-driven VS Code performance or memory investigations. Use when asked to launch Code OSS, automate a VS Code scenario, run the Chat memory smoke runner, capture renderer heap snapshots, take workflow screenshots, compare run summaries, or drive a repeatable scenario before heap-snapshot analysis.
chat-perf
Run chat perf benchmarks and memory leak checks against the local dev build or any published VS Code version. Use when investigating chat rendering regressions, validating perf-sensitive changes to chat UI, or checking for memory leaks in the chat response pipeline.
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…