Getting it into your agent
This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.
/plugin marketplace add LarryStanley/pi-delegate/plugin install pi-delegateWrote 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/larrystanley/pi-delegate/delegating-to-pi)<a href="https://agentmods.dev/skills/larrystanley/pi-delegate/delegating-to-pi"><img src="https://agentmods.dev/badge/skills/larrystanley/pi-delegate/delegating-to-pi/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/skills/larrystanley/pi-delegate/delegating-to-pi"><img src="https://agentmods.dev/badge/skills/larrystanley/pi-delegate/delegating-to-pi.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 | $0.00063 | $0.02664 |
| Opus 5 | $0.00032 | $0.01332 |
| Sonnet 5 | $0.00013 | $0.00533 |
| Haiku 4.5 | $0.00006 | $0.00266 |
Grade A, and why
delegating-to-pi 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 9d 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 — 168 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Delegating to pi
You are the tech lead, not the typist. All source code is written by pi — implementation and tests alike.
You produce only four things: the probe, the task book, the acceptance script, and the verdict. Every other character gets dispatched.
The test is structural, not a judgment call: is this a character that gets committed? Then dispatch it.
The four-way split (the one decision that must be right)
| What this really is | Who does it | The test |
|---|---|---|
| Writing any source code: implementation, tests, patches | pi | "Is this a character that gets committed?" |
| A lookup, decidable without reading context | a script | "Does this transformation require understanding context?" |
| Comparing two lists item by item | pi | Extract and normalize with a script, hand the comparison to pi |
| Deciding the contract, judging success/failure | you | probe, task book, acceptance script, verdict |
The "a lookup, decidable" branch is the most expensive misjudgment, and no hook can catch it. Measured: moving 33 files' declarations took a 46-minute dispatch + 5 timeouts + drove the endpoint down to 14.6 tok/s (normally 50), while a deterministic script finished in seconds at zero load. It needs no context understanding, so it should never pass through a model at all.
"pi can't handle this" is not a reason to do it yourself. Sometimes it means split the work — but reach for that far less often than feels natural, because splitting has a measured cost of its own (see below). The one place the 700-line rule genuinely applies is editing a single existing file: past roughly 700 lines pi spends its budget re-reading. Extract the part that needs changing into a new file and hand pi the whole new file. That is a rule about one file's size, not about how much work a dispatch may carry.
Size the task like a subagent's, not like a slice
Give pi a whole coherent unit of work. Multiple files, implementation and tests together, the whole feature. Do not pre-slice it into pieces small enough to feel safe.
What ships with it
7 files 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.
- 9d ago First seen · 168 lines · 63 tokens per session scan A 6b074c289d9a
delegating-to-pi is a skill published in the GitHub repository LarryStanley/pi-delegate (2 stars, last pushed 16d ago), licensed MIT. It adds 63 tokens to every session and 2,664 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
nft-standards
Implement NFT standards (ERC-721, ERC-1155) with proper metadata handling, minting strategies, and marketplace integration. Use when creating NFT contracts, building NFT marketplaces, or implementing digital asset systems.
postgresql-table-design
Use this skill when designing or reviewing a PostgreSQL-specific schema. Covers best-practices, data types, indexing, constraints, performance patterns, and advanced features.
istio-traffic-management
Configure Istio traffic management including routing, load balancing, circuit breakers, and canary deployments. Use when implementing service mesh traffic policies, progressive delivery, or resilience patterns.
event-store-design
Design and implement event stores for event-sourced systems. Use when building event sourcing infrastructure, choosing event store technologies, or implementing event persistence patterns.
projection-patterns
Build read models and projections from event streams. Use when implementing CQRS read sides, building materialized views, or optimizing query performance in event-sourced systems.
workflow-orchestration-patterns
Design durable workflows with Temporal for distributed systems. Covers workflow vs activity separation, saga patterns, state management, and determinism constraints. Use when building long-running processes, distributed transactions, or microservice orchestration.