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/understudylabs/understudy-agent-tools/product-knowledgenpx skills add understudylabs/understudy-agent-tools --skill product-knowledgegit clone --depth 1 https://github.com/understudylabs/understudy-agent-toolsWhat 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 | $0.00058 | $0.01123 |
| Opus 5 | $0.00029 | $0.00562 |
| Sonnet 5 | $0.00012 | $0.00225 |
| Haiku 4.5 | $0.00006 | $0.00112 |
Grade A, and why
product-knowledge 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 — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Product Knowledge
Explain Understudy as backend-agnostic infrastructure for improving LLM systems from real work: capture traces, run evals, compare model routes, optimize prompts or policies, and promote the route that best meets the developer's objective across quality, reliability, latency, cost, and constraints.
Use this skill for product explanations, onboarding copy, agent-facing help, and UI feature descriptions. Keep the answer concrete and tied to product surfaces the user can inspect.
For company identity, history, and the durable product narrative, read
reference.md. Keep the always-on explanation compact; load this
skill when the user needs the fuller story.
Safety Gates
Understudy defaults to the strongest active model and managed cloud execution unless the user selects Local or names a hard constraint. Dropping a workload starts analysis; launching a displayed plan authorizes its bounded upload, provider calls, hosted work, evaluation, receipts, and cleanup. State that envelope accurately and do not invent hidden destinations or spend.
Do not invent product availability. If a capability depends on local runtimes, warm model slots, account credentials, or a release channel, say that directly.
Resolve CLI
Product explanations usually do not need the CLI. When you need live local status,
prefer the installed understudy binary:
understudy status --json
If working inside a checkout where the binary is unavailable:
npm run build
node dist/bin.js status --json
Product Surfaces
- Desktop app — local control plane for chat, model serving, traces, evals, usage, account setup, and training workflows.
- Local serving — warm MLX slots for Understudy-suffixed local models, with first-run bootstrap for runtimes and model downloads.
- Chat harness — the Understudy agent runtime streams answers, reasoning, guarded tool calls, tool results, and compaction evidence to the native UI.
- Fusion sidekick — a smaller local model lane used for bounded read-only work while the main lane keeps planning, ambiguity, and final review.
- Evals / rollout lab — run task suites across model candidates and harness modes, watch each rollout, persist scores, and inspect failures.
- Candidate results — Test Results-style view that groups model-family task outcomes into passed, failed, running, skipped, score, latency, and drilldown rows.
- Training — progression from evals to GEPA/prompt optimization, datasets, SFT, RL, and distributed rollout jobs.
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.
- 2d ago First seen · 103 lines · 58 tokens per session scan A ea0c4593d0ad
product-knowledge is a skill published in the GitHub repository understudylabs/understudy-agent-tools (16 stars, last pushed 3d ago), licensed MIT. It adds 58 tokens to every session and 1,123 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 2026-08-30.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
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…
babysit-pr
Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…
imagegen
Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…
agent-host-chat-contributions
Build and review cross-cutting agent-host chat behavior through lifecycle contributions. Use when adding turn lifecycle side effects, prompt or context injection, restored-history transformation, protocol-action observation, or when reviewing changes that add code to AgentSideEffects or AgentService.