// Create mdflow-backed agent files for Script Kit. Compatibility path — prefer creating skills (SKILL.md under a plugin's skills/ directory) for new reusable AI work.
Use the local agy CLI as a fast Script Kit GPUI app inspector by prompting it to drive existing script-kit-devtools primitives, capture logs, and produce a compact investigation result from a user bug report or inspection prompt.
Manage the Notes window — creating, searching, editing, and automating notes via the SDK and automation protocol.
Canonical bootstrap for isolated Script Kit GPUI DevTools sessions: one front door script (classify, verify-script, start, prove, cleanup), repo-level helpers under scripts/agentic/, bounded timeouts, JSON stdout, progress stderr. Use when agents need parallel DevTools workers, isolated runtime proof, SK_VERIFY script gates, or dev-watch reuse without starting a second GPUI under ./dev.sh.
ACP Agent Chat lifecycle, AcpChatView, embedded/detached chat windows, agent/model selection, setup cards, streaming, cancellation, and chat close/reuse behavior.
ACP composer input, slash commands, @mentions, context parts, attachment tokens, pasted text/image, /new-script, focused target, and context staging.
Cmd+K, actions dialog, action sections, PromptPopup, confirm popup, attached popup, popup registry, popup resize/focus, and action execution routing.
| name | new-agent |
| description | Create mdflow-backed agent files for Script Kit. Compatibility path — prefer creating skills (SKILL.md under a plugin's skills/ directory) for new reusable AI work. |
Create mdflow-backed agent files for Script Kit. For new reusable AI work, prefer creating a skill (skills/<name>/SKILL.md) under the appropriate plugin instead — skills are first-class main-menu launchables that open Agent Chat.
~/.scriptkit/plugins/main/agents/*.md
| Filename | Meaning |
|---|---|
review.claude.md | Claude backend |
plan.agy.md | AGY backend |
code.codex.md | Codex backend |
triage.copilot.md | Copilot backend |
chat.i.agy.md | Interactive agent (.i. or _interactive: true) |
generic.md | Generic agent; use _command when needed |
Use underscore-prefixed keys so Script Kit metadata does not leak to backend CLI flags.
_sk_name_sk_description_sk_icon_sk_alias_sk_shortcut_inputs_interactive or _i_cwd_env_command---
_sk_name: "Review PR"
_sk_description: "Review staged changes and call out risks"
_sk_icon: "git-pull-request"
model: sonnet
---
Review the current git diff.
Return:
1. findings ordered by severity
2. concrete fixes
3. tests to add
---
_sk_name: "Plan Feature"
_sk_description: "Turn a feature request into an implementation plan"
_sk_icon: "map"
_interactive: true
_inputs:
feature_name:
type: text
message: "Feature name?"
risk_tolerance:
type: select
message: "Risk tolerance?"
choices: ["low", "medium", "high"]
model: gemini-2.0-flash
---
Create an implementation plan for {{ feature_name }}.
Risk tolerance: {{ risk_tolerance }}.
Include:
- files to change
- data flow
- tests to add
- rollback plan
scripts/export const metadatask_* keys