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/minmax/pi-cli-mcp/pi-contractnpx skills add minmax/pi-cli-mcp --skill pi-contractgit clone --depth 1 https://github.com/minmax/pi-cli-mcpWhat 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 | $0.00067 | $0.00914 |
| Opus 5 | $0.00034 | $0.00457 |
| Sonnet 5 | $0.00013 | $0.00183 |
| Haiku 4.5 | $0.00007 | $0.00091 |
Grade A, and why
pi-contract 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 — 74 lines — stays where its author put it; the contents beside it link to each section on GitHub.
pi's contract
Everything this server reads is pi's data: the --mode json event stream and the stopReason on
each assistant message. The failure mode is silent drift — pi adds a stop reason or renames an event,
the runtime quietly files it under "unknown", and a delegated answer comes back wrong or a good one
is reported as failed.
Types are borrowed, not re-described
src/types.ts imports pi's own types:
import type { ThinkingLevel as PiThinkingLevel } from "@earendil-works/pi-agent-core";
import type { AssistantMessage, StopReason, TextContent, ToolCall, Usage } from "@earendil-works/pi-ai";
import type { JsonAgentSessionEvent } from "@earendil-works/pi-coding-agent";
import type is erased at compile time, so those packages stay devDependencies and the published
package has zero dependencies. test/packaging.test.ts enforces both halves: every pi import must be
a type import, and nothing outside node: builtins may be imported at runtime.
Do not hand-copy a shape from pi into this repo. Alias it.
Stop reasons are a partition, checked both ways
pi's vocabulary is "pending" | "stop" | "length" | "toolUse" | "error" | "aborted" | "deferred".
src/types.ts splits it into four buckets with satisfies readonly PiStopReason[], which rejects a
reason pi does not have. test/pi-contract.test-d.ts checks the other direction — that the buckets
stay exhaustive and disjoint over pi's union — plus that the event names we branch on still exist and
that the usage fields we read keep their types.
Run it with npm test (typecheck is on) or npx vitest run --typecheck.only.
This is not theoretical. Writing that test immediately found:
pendinganddeferredwere unknown to the classifier, so an ordinary intermediate message would have failed the call;auto_compaction_start— a branch on an event pi does not have. The name came from an older unrelated fork; pi's iscompaction_start. Dead code, silently never matching.tool_execution_start.tool— a fallback field pi never sends.
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 · 74 lines · 67 tokens per session scan A e2bb59b201cb
pi-contract is a skill published in the GitHub repository minmax/pi-cli-mcp (0 stars, last pushed 10d ago), licensed MIT. It adds 67 tokens to every session and 914 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 2026-08-31.
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…