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 instructions/tentacleopera/switchboard/claude-mdgit clone --depth 1 https://github.com/TentacleOpera/switchboardWhat 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 | $0.05110 | $0.05110 |
| Opus 5 | $0.02555 | $0.02555 |
| Sonnet 5 | $0.01022 | $0.01022 |
| Haiku 4.5 | $0.00511 | $0.00511 |
Grade A, and why
switchboard CLAUDE.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 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.
This is a copy
86% identical to switchboard AGENTS.md — 89 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 224 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Switchboard — agent rules
NEVER add confirmation dialogs. NO EXCEPTIONS.
Delete buttons delete immediately. No confirm(), no window.confirm(), no modal showWarningMessage, no two-click patterns, no "Are you sure?". The user has demanded this repeatedly. Buttons are deliberately hard to misclick.
Also a hard technical reason: window.confirm() is a silent no-op in VS Code webviews (sandboxed iframe without allow-modals — it always returns false). Any confirm gate added to src/webview/planning.js, src/webview/kanban.html, etc. makes the button do literally nothing. This exact bug broke the kanban delete-plan button (fixed 2026-06-11).
If you find a confirm gate in this codebase, it is a bug — remove it. Multi-choice decision dialogs (e.g. 3-way conflict resolution) are allowed; plain confirm gates are not.
Standalone and the extension MUST NOT diverge. NO EXCEPTIONS.
Switchboard ships two hosts: the VS Code extension (src/extension.ts) and
the standalone/npx host (src/standalone/bootstrap.ts). Every feature must land
in both. If you are planning a feature, the plan names both composition
roots and its verification covers both. If you are implementing one, the diff
touches both. "Extension first, standalone later" is not a plan — it is a
divergence, and no gate catches it.
The trap is not verbs. bootstrap.ts's default: arm delegates every
unmatched verb to the provider, so verb-reachability audits always come back
green. The trap is composition-root wiring: service seams
(engine.setX(...)), options objects handed to shared services, and
Promise<void> callbacks where "never wired" and "working" are the same value.
Diff the two roots by hand. The seams each host wires are the audit — not the
verbs each host answers.
Precedent (2026-08). All four PlanIngestionEngine queue seams —
setQueueHeadResolver, setQueuePacingResolver, setQueueTeamMembersResolver,
setQueueEscalationRecorder — were wired in extension.ts only, a month after
standalone shipped. Consequently no queue watch was ever armed in the
standalone host: seat pacing was unreachable, dead seats were never re-staged,
and the queue stall backstop did not exist. Every gate stayed green, because
npm run standalone-parity:check is scoped to the browser read-back path, not
the composition root. The two roots had also drifted the other way — standalone
wires two seams the extension does not.
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 · 224 lines · 5,110 tokens per session scan A 0c03660dfdd7
switchboard CLAUDE.md is an instructions file published in the GitHub repository TentacleOpera/switchboard (213 stars, last pushed 3d ago), licensed MIT. It adds 5,110 tokens to every session, about $0.0255 per session on Opus 5. A static security scan graded it A with 0 findings. It is 86% identical to switchboard AGENTS.md, differing in 89 lines, and is treated as a copy.
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).
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
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 oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
next.js AGENTS.md
Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.