/brief - Artifact Brief Route
SkillAI & modelsUse when the user wants the current conversation or captured context turned into a lifecycle artifact without a long interview, especially before proposal planning.
Available today. Use it from your connected AI after setup.
No other account needed.
Connect ahel once, and every AI you use reads what you have installed.
Then ask your AI: use the /brief - Artifact Brief Route skill
What this skill tells your AI
The instructions your AI receives, as published by lev-os/agents in skills-db/_todo/brief/SKILL.md and read by ahel’s review.
/brief is a thin lifecycle route into interview --auto. It exists because
/design reads as UI/UX even when the needed output is a product,
architecture, decision, or planning artifact.
Work Link
Lifecycle lane: Shape -> Plan
Entity movement: captured | context -> artifact_ready | needs_interview | needs_prior_art
Workstream: resolve active workstream before writing artifacts
Upstream: /capture, /dump, conversation, /work
Downstream: /interview --auto, /prior-art, /propose
Router: /work
HUD: end with 🧬 {ws} ⚡{exec_count} 📥{capture_count} ⏸️{paused_count} ✅{done_count} | 🚦brief={gate}/{score} | ⏭️ {next} | 🔁{loop_state}
Commands
/brief # run /interview --auto --standard --compact
/brief --quick # auto artifact pass, ambiguity gate <= 0.30
/brief --standard # default, ambiguity gate <= 0.20
/brief --deep # higher rigor, ambiguity gate <= 0.15
/brief --full # show richer evidence and trade-offs
/capture --brief # capture delegation route into /brief
Protocol
steps:
- id: resolve_context
action: Identify the active workstream, source conversation, and target artifact subject.
validation: "Workstream, subject, source context, and likely artifact path are known."
on_failure: "Route to /capture or /ws before writing."
- id: choose_depth
action: Select quick, standard, or deep; default to standard.
validation: "Depth threshold is explicit."
on_failure: "Use --standard."
- id: run_auto_interview
action: Delegate to /interview --auto with the selected depth and output mode.
validation: "The auto pass inspects context before asking."
on_failure: "Ask one blocking ambiguity question."
- id: write_or_route
action: Write/update the lifecycle artifact or route to prior-art/interview.
validation: "Artifact path, blocker, or next owner is explicit."
on_failure: "Keep the item captured; do not promote to /propose."
- id: offer_next
action: Route artifact-ready output to /propose or keep it in capture.
validation: "Next lifecycle step is one of propose, prior-art, interview, capture, or hold."
on_failure: "Emit the HUD with needs_interview."
Artifact Target
For now, /brief still writes through the existing interview artifact contract:
.lev/pm/designs/<slug>.md using the work design template. The surface name is
brief; the storage contract stays unchanged until decisions, PM mirroring, and
cascading contexts are aligned.
Auto Gate
auto_gate:
pass_when:
- ambiguity <= selected_depth_threshold
- subject is named
- source context is cited
- artifact owner/path is known
- unresolved human judgments are not blocking
ask_when:
- subject is unclear
- artifact path would be invented
- constraints or success criteria are below threshold
- prior-art is required before safe synthesis
Output Contract
brief_result:
subject: ""
source_context: []
depth: quick|standard|deep
ambiguity: 0.0
artifact_path: ""
status: artifact_ready|needs_interview|needs_prior_art|captured
next: propose|interview|prior-art|capture|hold
Red Baseline
- Do not use
/designas the user-facing command for this route. - Do not turn
/briefinto a UI/UX design skill. - Do not write an artifact when ambiguity is above the selected gate.
- Do not promote to
/proposewithout a concrete artifact path or captured blocker.
Related
/capture --briefdelegates here./interview --autodoes the artifact synthesis./proposeconsumes aligned artifact output.
Signals
- GitHub stars
- 22
- Forks
- 2
- Last commit
- Sep 2026
Advanced
- Catalog kind
- skill
- Gateway key
brief-lev-os- Source
- github.com/lev-os/agents