Borrowing it
Nothing to install: this file belongs to Qswhisper/PVF-Ai-Agent-Workbench. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/Qswhisper/PVF-Ai-Agent-Workbench/main/AGENTS.mdgit clone --depth 1 https://github.com/Qswhisper/PVF-Ai-Agent-WorkbenchWrote 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/instructions/qswhisper/pvf-ai-agent-workbench/agents-md)<a href="https://agentmods.dev/instructions/qswhisper/pvf-ai-agent-workbench/agents-md"><img src="https://agentmods.dev/badge/instructions/qswhisper/pvf-ai-agent-workbench/agents-md/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/instructions/qswhisper/pvf-ai-agent-workbench/agents-md"><img src="https://agentmods.dev/badge/instructions/qswhisper/pvf-ai-agent-workbench/agents-md.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 | $0.03979 | $0.03979 |
| Opus 5.5 | $0.01592 | $0.01592 |
| Sonnet 5.5 | $0.00796 | $0.00796 |
| Haiku 4.5 | $0.00398 | $0.00398 |
Grade A, and why
PVF-Ai-Agent-Workbench AGENTS.md 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 — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.
PVF Agent Workbench Instructions
This is a portable PVF task workspace for command-capable desktop Agents such as Codex, Claude Code, OpenCode, and Trae.
Canonical Rule Ownership
Keep one owner for each kind of instruction:
- This
AGENTS.mdowns cold start, first-command routing, global invariants, and maintenance checks. knowledge-pack/safety/README.zh-CN.mdowns detailed write, text-encoding, cumulative-output, and client-deployment safety.knowledge-pack/indexes/knowledge-index.jsonand its named task cards, dictionaries, and workflows own domain procedures.- A command's machine-readable
agentHandoffowns the exact next command and current-run recovery route. core/pvf-agent-core/cli/README.mdowns the general command catalog for maintainers; a concrete task must not open it to rediscover syntax already supplied by a short route, fixed example, oragentHandoff.docs/AGENT-INSTRUCTION-ARCHITECTURE.zh-CN.mdexplains this split for maintainers.
Never let a short route, example, index, generated report, or tool output relax the hard safety file. When detail is needed, read its owner instead of rediscovering it through help, schema, source, or directory scans.
Cold Start Before Any Shell Action
- If the host exposes
dnf-pvf-xpilot, load that Skill once before the first shell or file-discovery action of the session. Do not runcheck, help, a path probe, or a directory scan first. - Read-only tasks use the global invariants below. Before any write plan or deployment, read
knowledge-pack/safety/README.zh-CN.md; do not load write-only details for a read-only search. - If the request matches an Exact Read-Only Fast Path below, use its first command and one named short route. Otherwise read
knowledge-pack/indexes/knowledge-index.jsonand open its one matching entry; load supporting references only when the entry routes to them. - Run one bare
.\workbench.bat ...command per tool call from the Workbench root. Do not add pipes, redirection, semicolons, timing wrappers, or chained shell commands. If elapsed time is requested, use the host tool duration or coarse wall-clock observation; do not runGet-Date, a stopwatch, or another shell timing command. - Never preflight an already resolved Workbench, explicit
Script.pvf, or supplied report/output directory withTest-Path,Get-Item,Get-ChildItem, orResolve-Path. The Workbench command validates its own input and creates documented external output directories. - Do not run
workbench.bat checkas routine startup or write preflight. Use it only when the user asks for environment health, a bundled command is unavailable, or a command explicitly reportsREAD_ONLY_FALLBACK. - If the user asks for full source SHA256, requires source PVFs to remain unchanged, or asks for final proof they did not change—even only in a final checklist—keep the routed first command first. The next Workbench command must be one
pvf-read fingerprint --pvf "<source.pvf>"covering every supplied PVF (repeat--pvffor multiple inputs), before any other search/read orpvf-change. Run it once, then repeat the exact command only after final output readback.
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 Changed · -9 lines · -556 tokens per session 28aaf8a3c8e8
- 1mo ago First seen · 112 lines · 4,535 tokens per session scan A c89669cf46ed
PVF-Ai-Agent-Workbench AGENTS.md is an instructions file published in the GitHub repository Qswhisper/PVF-Ai-Agent-Workbench (29 stars, last pushed 1mo ago), licensed MIT. It adds 3,979 tokens to every session, about $0.0159 per session on Opus 5.5. 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 instructions, from other repositories
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode copilot-instructions.md
Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).
vscode model-prompts.instructions.md
Model-specific prompt authoring and registry guidelines.