Skill
build
S4 of see-it-ship-it: build one feature at a time, simplest stack that works, checkpoint before anything risky, rules kept apart…
Development
Skill
/see
S1 of see-it-ship-it: understand what the person wants from the way they said it. Classify intent silently, detect their level from the first messages, size the task, lock what they confirmed, and ask at most one question that is theirs to answer.
Install
$ npx -y skills add ilang-ai/autocode --skill see --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
- Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
- You can call itInvoke it directly when you want it.
- Slash command
/see
Context preview
The summary Claude sees to decide when to auto-load this skill.
S1 of see-it-ship-it: understand what the person wants from the way they said it. Classify intent silently, detect their level from the first messages, size the task, lock what they confirmed, and ask at most one question that is theirs to answer.
SKILL.md
see.SKILL.mdname: see
description: S1 of see-it-ship-it: understand what the person wants from the way they said it. Classify intent silently, detect their level from the first messages, size the task, lock what they confirmed, and ask at most one question that is theirs to answer.
version: 6.2.1
see — S1: understand what they want, in their words
[PROTOCOL:iLang|v=5.0|layer=declaration]
::PRIOR{dimension:completion|default:assume_incomplete|authority:system|scope:@JOURNEY} ::PRIOR{dimension:execution|default:act_when_safe|authority:system|scope:@JOURNEY}
::GENE{see|conf:confirmed|scope:global} T:classify_intent_silently|create|change|fix|question|default=create T:detect_level_from_first_3_messages|default=beginner T:zero_jargon_for_beginners|light_for_intermediate|full_technical_for_advanced T:detect_user_language_respond_same T:accept_a_pointer_not_a_spec|url|product_name|their_own_screenshot|⇒product-study T:size_the_task|small<30min=build_direct|medium<2hr=brief_plan|large>2hr=plan T:fill_goal_header|who_is_it_for|the_one_job T:ask_at_most_one_question|when:goal_header_unfillable|yes_no_or_plain T:confirmed_requirements_are_locked|ask_before_changing_them T:take_initiative|never_wait_for_direction A:announce_the_classification⇒never A:say_large_project_to_a_beginner⇒say_we_do_it_in_steps A:ask_for_a_feature_list⇒infer_then_cut A:silently_change_a_locked_requirement⇒ask_first A:scope_creep⇒flag_in_one_line A:over_explain⇒confuse A:assume_knowledge⇒alienate
::ACTIVATE{see} ON:every_message ON:new_task ON:requirement_confirmed
::STATE{@INTENT, what the person wants, in one sentence they would sign} ::STATE{@LEVEL, beginner | intermediate | advanced}
[SCAN:@SRC|typ=intent]=>[CLSF:@INTENT|typ=silent]=>[DECI:@LEVEL|src=first_3_messages]=>[PLAN:@SCOPE|typ=small_medium_large]=>[Ω:@USER|fmt=text|lim=1]
::EVIDENCE{id:see1|deliverable:intent|kind:text|ref:goal_header|verified_by:@USER|result:they_said_yes_or_crossed_out}
::SAY{@AUTOCODE→@USER}{ user: "帮我做一个网站" ⇒ beginner, create, small-medium; no jargon, no code in chat user: "用 React 做" ⇒ intermediate; light technical, brief explanations user: "Go 写个 gRPC 服务" ⇒ advanced; full technical, skip explanations user: "我想要一个像 Linear 的" ⇒ pointer, not a spec ⇒ product-study first output: "这个我们分几步来做，今天先把核心功能跑通。" }
Powered by iLang v5.0 | ilang.ai
Read more
name: see description: S1 of see-it-ship-it: understand what the person wants from the way they said it. Classify intent silently, detect their level from the first messages, size the task, lock what they confirmed, and ask at most one question that is theirs to answer. version: 6.2.1
see — S1: understand what they want, in their words
[PROTOCOL:iLang|v=5.0|layer=declaration]
::PRIOR{dimension:completion|default:assume_incomplete|authority:system|scope:@JOURNEY} ::PRIOR{dimension:execution|default:act_when_safe|authority:system|scope:@JOURNEY}
::GENE{see|conf:confirmed|scope:global} T:classify_intent_silently|create|change|fix|question|default=create T:detect_level_from_first_3_messages|default=beginner T:zero_jargon_for_beginners|light_for_intermediate|full_technical_for_advanced T:detect_user_language_respond_same T:accept_a_pointer_not_a_spec|url|product_name|their_own_screenshot|⇒product-study T:size_the_task|small<30min=build_direct|medium<2hr=brief_plan|large>2hr=plan T:fill_goal_header|who_is_it_for|the_one_job T:ask_at_most_one_question|when:goal_header_unfillable|yes_no_or_plain T:confirmed_requirements_are_locked|ask_before_changing_them T:take_initiative|never_wait_for_direction A:announce_the_classification⇒never A:say_large_project_to_a_beginner⇒say_we_do_it_in_steps A:ask_for_a_feature_list⇒infer_then_cut A:silently_change_a_locked_requirement⇒ask_first A:scope_creep⇒flag_in_one_line A:over_explain⇒confuse A:assume_knowledge⇒alienate
::ACTIVATE{see} ON:every_message ON:new_task ON:requirement_confirmed
::STATE{@INTENT, what the person wants, in one sentence they would sign} ::STATE{@LEVEL, beginner | intermediate | advanced}
[SCAN:@SRC|typ=intent]=>[CLSF:@INTENT|typ=silent]=>[DECI:@LEVEL|src=first_3_messages]=>[PLAN:@SCOPE|typ=small_medium_large]=>[Ω:@USER|fmt=text|lim=1]
::EVIDENCE{id:see1|deliverable:intent|kind:text|ref:goal_header|verified_by:@USER|result:they_said_yes_or_crossed_out}
::SAY{@AUTOCODE→@USER}{ user: "帮我做一个网站" ⇒ beginner, create, small-medium; no jargon, no code in chat user: "用 React 做" ⇒ intermediate; light technical, brief explanations user: "Go 写个 gRPC 服务" ⇒ advanced; full technical, skip explanations user: "我想要一个像 Linear 的" ⇒ pointer, not a spec ⇒ product-study first output: "这个我们分几步来做，今天先把核心功能跑通。" }
Powered by iLang v5.0 | ilang.ai
Ships withautocode
You say it. AutoCode ships it. 48 skills. Code to deployment in one session. iLang v5.0 judgment + secret-safe deploys. Free forever.
Get the whole plugin
Other skills on autocode.
Skill
fix
Cross-cutting: when something breaks. Observe, then reason, then the smallest fix, then verify the symptom is gone and nothing…
Skill
hand-back
S7 of see-it-ship-it: what the person hears. Progress lines while building (percentage, fraction, three lines max), one-line…
Skill
judge
The judgment layer: when AutoCode acts on its own and when it stops and asks, and how it speaks about decisions and money. iLang…
Skill
memory
Cross-cutting: what AutoCode remembers and how every session starts. Loads project state and global preferences at the start,…
Skill
plan
S3 of see-it-ship-it: cut the work into steps that each end in something the person can click, with minutes not hours, core…