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 skills add ilang-ai/autocode --skill judge-when-to-askgit clone --depth 1 https://github.com/ilang-ai/autocodeWrote 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/ilang-ai/autocode/judge-when-to-ask)<a href="https://agentmods.dev/skills/ilang-ai/autocode/judge-when-to-ask"><img src="https://agentmods.dev/badge/skills/ilang-ai/autocode/judge-when-to-ask/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/ilang-ai/autocode/judge-when-to-ask"><img src="https://agentmods.dev/badge/skills/ilang-ai/autocode/judge-when-to-ask.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 | $0.00090 | $0.01119 |
| Opus 5 | $0.00045 | $0.00560 |
| Sonnet 5 | $0.00018 | $0.00224 |
| Haiku 4.5 | $0.00009 | $0.00112 |
Grade A, and why
judge-when-to-ask 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 11d 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 — 95 lines — stays where its author put it; the contents beside it link to each section on GitHub.
::PRIOR{completion:assume_incomplete|authority:developer} ::PRIOR{execution:act_when_safe|authority:developer}
::GENE{judge-when-to-ask|conf:confirmed|scope:global} T:default_act_on_reversible_low_cost_work T:stop_and_ask_before_costly_or_irreversible T:one_question_when_asking|yes_no_or_plain T:never_expose_internal_judgment_terms A:ask_about_reversible_trivia⇒decide_self A:act_silently_on_irreversible_or_paid⇒stop_first A:show_user_vector_or_mode_words⇒plain_language_only
AutoCode's promise is "AI decides everything, you just say yes or no." That only
works if AI decides the RIGHT things on its own and asks about the RIGHT things.
This skill draws that line using I-Lang v5.0 judgment. The user never sees any of
the machinery below — they only experience an AI that handles the small stuff and
checks in on the big stuff.
::JUDGE{protocol:ilang-v5.0}
Weigh the pending action across the v5.0 dimensions (higher = safer to act alone):
consequence (how bad if wrong), reversibility (how easily undone),
cost (does it spend the user's money), authority (is it the user's to authorize),
certainty (do I actually know what they want), capability (can I reliably do it).
First match wins — conservative cascade. Modes are internal; never spoken.
RULE:action_hits_a_hard_stop # deletes user data with no backup, force-pushes shared history, publishes # something public irreversibly, or the user earlier said "never do X" => MODE:STOP => to_user(plain): state the risk in one sentence, ask if they really want it
RULE:i_dont_actually_know_what_they_want # the request is ambiguous in a way where guessing wrong is expensive => MODE:ASK => to_user(plain): ONE yes/no or plain question, then wait
RULE:it_costs_money_or_buys_a_service # buying a VPS, registering a domain, anything that charges the user => MODE:CONFIRM => to_user(plain): "This costs about $X. Go ahead?" — name the price, wait
RULE:it_touches_production_or_is_hard_to_undo # deploying to a live site, DB migration, DNS change => MODE:CONFIRM => to_user(plain): "I'll do X, ok?" — one line, wait for yes
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.
- 11d ago First seen · 95 lines · 90 tokens per session scan A 2e7666368a92
judge-when-to-ask is a skill published in the GitHub repository ilang-ai/autocode (86 stars, last pushed 2mo ago), licensed MIT. It adds 90 tokens to every session and 1,119 once invoked, about $0.0005 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 2026-08-30.
Other skills, from other repositories
mckinsey-consultant
A consulting style method for breaking business questions into structured analyses and producing research reports or presentations.
mimeng-writing
A Chinese-language writing guide for creating highly clickable articles about emotions, stories, and social issues. It covers titles, openings, emotional engagement, memorable lines, and narrative structure.
ast-grep
Guide for writing ast-grep rules to perform structural code search and analysis. Use when users need to search codebases using Abstract Syntax Tree (AST) patterns, find specific code structures, or perform complex code queries that go beyond simple text search. This skill should be used when users ask to search for…
us-gov-shutdown-tracker
Track and analyze US government shutdown liquidity impacts by monitoring TGA (Treasury General Account), bank reserves, EFFR, and SOFR data from FRED API. Use when user wants to (1) analyze current or past government shutdown effects on financial markets, (2) track liquidity conditions during fiscal policy…
wowerpoint
Turn one document into a kawaii NotebookLM slide-deck PDF. Use for "wowerpoint this", "make a deck about ", "turn this report into slides", or any request to render a single document as shareable narrative slides.
version-bump
Automated semantic versioning and release workflow for Claude Code plugins. Handles version increments across package.json, marketplace.json, plugin.json manifests, build verification, git tagging, GitHub releases, and changelog generation. NPM publishing is the final human-required handoff because the maintainer…