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/perhapsspy/project-legibility/interactive-state-flownpx skills add perhapsspy/project-legibility --skill interactive-state-flowgit clone --depth 1 https://github.com/perhapsspy/project-legibilityWhat 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 | $0.00065 | $0.01170 |
| Opus 5 | $0.00032 | $0.00585 |
| Sonnet 5 | $0.00013 | $0.00234 |
| Haiku 4.5 | $0.00006 | $0.00117 |
Grade A, and why
interactive-state-flow 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 — 73 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Interactive State Flow
Purpose
Keep interactive software responsive and correct when user intent, expensive presentation, and async work progress at different speeds.
Use this skill for typing, selection, navigation, search, previews, streaming, realtime updates, file or log views, and similar flows when there is observable lag, stale-result risk, or mixed responsibility. Do not introduce extra state or scheduling boundaries into small, clear, responsive code merely because it uses async work, caching, or effects.
Runtime Contract
Keep source truth prompt
Promptly record the interaction-owned intent or input state, and commit authoritative source state through its owner, without waiting for expensive derivation or presentation. This does not require synchronous rendering or premature durable mutation; persistence remains governed by its own authority and ordering contract. Presentation state includes filtered rows, previews, rendered ranges, and pending, cached, stale, or progressive output.
Presentation may lag only when that preserves responsiveness and does not misrepresent current truth. Do not debounce the state the UI contract says is current merely to reduce downstream cost; defer its expensive consequences instead. Delayed commitment is valid when it is itself the declared product behavior.
Protect the urgent path proportionally
Name the user-visible contract before choosing a technique:
- what must update promptly
- what may lag, be skipped, or remain stale
- which context makes a result obsolete
- which boundary decides whether a result may commit
Classify follow-up work by responsibility, urgency, cost, visibility, and freshness risk. Then use the simplest mechanism that satisfies the user-visible contract.
Do not start from a preferred mechanism such as debounce, memoization, a transition, or a worker. Moving work off-thread or into another process is justified only when it protects the interaction path and the boundary has clear input, output, ownership, ordering, cancellation, and failure behavior. Account for transfer and coordination cost. Background work returns candidates; it must not bypass the accepting owner to mutate current UI or presentation.
What ships with it
2 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.
- 2d ago First seen · 73 lines · 65 tokens per session scan A 06042b384888
interactive-state-flow is a skill published in the GitHub repository perhapsspy/project-legibility (5 stars, last pushed 5d ago), licensed MIT. It adds 65 tokens to every session and 1,170 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
author
Write a new doc in the right place with the right name. Use when asked to "write a doc", "where should this doc go", "add an ADR", "document this decision", "new spec", or "new feature plan".
router
Find the right doc before reading code. Use when asked "where is X documented", "is there a spec/ADR for X", "what's the roadmap", "open bugs", "read the docs about X", or "share the docs".
changelog
Record shipped work in the monthly changelog (append-only). Use when something ships or when asked "what changed", "add to changelog", "release notes", or "what shipped this month".
init
Set up docflow in a repo with no docs yet: scaffold the docs tree, write docflow.json, add agent guidance. Use when asked to "set up docflow", "initialize docs", "scaffold docs", or "make this repo use docflow".
check
Friendly docflow readiness check: one status, one reason, and the exact next command. Use when asked if docflow is set up, ready, or "what do I do next".
doctor
Read-only docflow diagnosis: scans docs, config, changelog, and links, then recommends init, adopt, or repair. Use when asked "check docs setup", "should I set up docflow", "why is docflow not working", or "doctor".