@lifeaitools/rdc-skills 0.9.24 → 0.9.26

package/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "rdc",
-  "version": "0.9.24",
+  "version": "0.9.25",
   "description": "RDC typed-agent dispatch skill suite for Claude Code — plan, build, review, overnight unattended builds with work-item tracking and TDD enforcement.",
   "author": {
     "name": "LIFEAI",

package/README.md

@@ -176,7 +176,7 @@ python3 -c "import json; print(json.load(open('package.json'))['version'])"
 /rdc:help
 ```
 
-Current version: **v0.9.24**
+Current version: **v0.9.25**
 
 ## Quick Start
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@lifeaitools/rdc-skills",
-  "version": "0.9.24",
+  "version": "0.9.26",
   "description": "RDC typed-agent dispatch skill suite for Claude Code — plan, build, review, overnight builds",
   "keywords": [
     "claude-code",
@@ -19,7 +19,7 @@
     "type": "plugin",
     "skills": "skills/",
     "guides": "guides/",
-    "version": "0.9.24",
+    "version": "0.9.26",
     "commands": "commands/"
   },
   "scripts": {

package/skills/design/SKILL.md

@@ -27,6 +27,7 @@ RDC-owned design execution for Studio and LIFEAI interfaces. This skill is the S
 ## Arguments
 
 - `rdc:design` — show the command menu
+- `rdc:design edit <target>` — start a local source-edit session for a URL, app, brand, route, or file
 - `rdc:design studio <target>` — Studio-aware design task
 - `rdc:design tokens <brand-or-route>` — token model, resolver, export, or governance task
 - `rdc:design palette <brief>` — Palette Library or palette generation task
@@ -57,8 +58,10 @@ Project docs to read for Studio work:
 ## Procedure
 
 1. **Classify the task.**
+   - `edit` means start a local source-edit session. Do not show the generic command menu for `edit`.
    - `studio`, `tokens`, `palette`, and `theme` always load the Studio model.
    - `palette`, `theme`, and `colorize` also load the Rampa reference.
+   - Any task involving colors, palettes, ramps, contrast, neutrals, light/dark color pairs, status colors, chart colors, or theme colors loads the Rampa reference.
    - `audit`, `critique`, `polish`, and `craft` load Studio model when the target is inside Studio or uses Studio tokens.
 
 2. **Read the actual system context.**
@@ -72,24 +75,96 @@ Project docs to read for Studio work:
    - Live editor: RDC Native Local Debug only — `/editor/local/*`, `/api/editor/local-debug`, `/studio/debug/*`, `/studio/claude/*`, and `scripts/studio-debug-poll.mjs`.
    - Components: `@regen/ui`, Studio component registry, and existing local components.
 
-4. **Use Rampa only as proposal tooling.**
+4. **For `edit`, start the local session directly.**
+   - Intent syntax:
+     ```powershell
+     rdc:design edit <url|domain|brand|app|file>
+     ```
+   - Required variables:
+     - `target`
+     - `brandSlug`
+     - `appSlug`
+     - `repoRoot`
+     - `cwd`
+     - `connectorBaseUrl`
+     - `studioOrigin`
+     - `modeDefault`
+   - Fill missing variables from the obvious target mapping before asking:
+     - `test`, `studio_test`, `studio-test` -> `--brand test --app studio_test`
+     - `prt`, `prt-portal`, `prtrust.fund`, `dev.prtrust.fund` -> `--brand prt --app prt`
+   - If the target is a file path and brand/app are not obvious, ask one concise question for the missing brand/app. Do not search broadly or show the generic command menu.
+   - Canonical start call:
+     ```powershell
+     curl.exe -s -X POST "http://127.0.0.1:52437/studio/debug/start" `
+       -H "Content-Type: application/json" `
+       -d "{\"brandSlug\":\"<brandSlug>\",\"appSlug\":\"<appSlug>\",\"repoRoot\":\"C:/Dev/regen-root\",\"cwd\":\"C:/Dev/regen-root\",\"modeDefault\":\"direct_edit\",\"startedBy\":\"claude-cli\"}"
+     ```
+   - Expected successful response shape:
+     ```json
+     {
+       "ok": true,
+       "sessionId": "studio-...",
+       "token": "...",
+       "devUrl": "http://localhost:3006",
+       "relayBaseUrl": "http://127.0.0.1:52437/studio/debug/studio-...",
+       "claudeBaseUrl": "http://127.0.0.1:52437/studio/claude/studio-...",
+       "pollCommand": "node scripts/studio-debug-poll.mjs --session ... --token ... --relay ..."
+     }
+     ```
+   - Construct the Studio attach URL exactly:
+     ```txt
+     http://localhost:3011/editor/local/<brandSlug>?attach=1&sessionId=<sessionId>&token=<token>
+     ```
+   - Never send the user to `/editor/local/<brandSlug>` without `attach=1&sessionId=...&token=...`.
+   - Canonical poll call:
+     ```powershell
+     curl.exe -s "http://127.0.0.1:52437/studio/claude/<sessionId>/poll?token=<token>&timeout=600000"
+     ```
+   - Canonical reply call:
+     ```powershell
+     curl.exe -s -X POST "http://127.0.0.1:52437/studio/claude/<sessionId>/reply" `
+       -H "Content-Type: application/json" `
+       -d "{\"token\":\"<token>\",\"eventId\":\"<eventId>\",\"status\":\"done\",\"message\":\"Edited <path>\",\"filesChanged\":[\"<path>\"]}"
+     ```
+   - Allowed terminal statuses are `done`, `error`, `blocked`, and `needs_reference`.
+   - Convenience helper, equivalent to the canonical start call plus attach URL construction:
+     ```powershell
+     node scripts/studio-debug-start.mjs --target <url|domain|brand|app|file> [--brand <slug>] [--app <slug>] [--mode direct_edit|variant_edit|reference_replace|note] [--studio-origin http://localhost:3011] [--connector http://127.0.0.1:52437] [--repo-root C:/Dev/regen-root] [--cwd C:/Dev/regen-root] [--dev-url <url>] [--dev-command <command>] [--no-launch]
+     ```
+   - Helper output contract:
+     ```txt
+     STUDIO_URL=http://localhost:3011/editor/local/<brandSlug>?attach=1&sessionId=...&token=...
+     POLL_COMMAND=node scripts/studio-debug-poll.mjs --session ... --token ... --relay ...
+     SESSION_FILE=.studio-debug/<sessionId>.json
+     ```
+   - If using the helper, run:
+     ```powershell
+     node scripts/studio-debug-start.mjs --target <target>
+     ```
+   - Return `STUDIO_URL` and `POLL_COMMAND` to the user. If operating as Claude CLI, open or instruct opening `STUDIO_URL`, then run `POLL_COMMAND`.
+   - After an event is received, edit the source directly and reply with the helper or canonical reply call:
+     ```powershell
+     node scripts/studio-debug-poll.mjs --session <sessionId> --token <token> --relay <claudeBaseUrl> --reply <eventId> done --file <path> --message "Edited <path>"
+     ```
+
+5. **Use Rampa only as proposal tooling.**
    - Run Rampa CLI for color ramps, APCA/WCAG checks, tinted neutrals, status palettes, or data-viz palettes.
    - Treat output as draft design material.
    - Persist only after mapping into Studio token roles or Palette Library payloads.
 
-5. **Apply RDC design rules.**
+6. **Apply RDC design rules.**
    - Use Studio tokens for production surfaces.
    - Do not hardcode color/font/spacing when a token exists.
    - Do not modify locked governance tokens.
    - Keep product UIs compact, operational, and scannable.
    - Avoid generic AI design patterns and card-heavy layouts.
 
-6. **Plan edits before mutating files.**
+7. **Plan edits before mutating files.**
    - State the files/routes/tables involved.
    - Use RDC work-item protocol for non-trivial implementation.
    - Keep unrelated installed skills and vendor artifacts untouched.
 
-7. **Verify.**
+8. **Verify.**
    - Run scoped tests only.
    - For Studio: prefer route smoke checks, token API checks, and browser screenshots when UI changed.
    - For CLI prompt work: run `node C:/Dev/rdc-skills/scripts/rdc-design-cli.mjs <command> <brief>` and inspect the generated report under `.rdc/reports/rdc-design-cli/`.
@@ -98,6 +173,7 @@ Project docs to read for Studio work:
 
 | Command | Purpose |
 |---|---|
+| `edit` | Start a local source-edit session for a URL, app, brand, route, or file |
 | `studio` | Studio-aware design/build task with token, route, and editor context |
 | `tokens` | Diagnose or design token usage, resolver flow, API/export, governance |
 | `palette` | Palette Library work, palette generation, external Palette Designer handoff |

package/skills/fixit/SKILL.md

@@ -93,7 +93,7 @@ Submit implementation report first, then mark done:
 
 ```sql
 SELECT submit_implementation_report('<id>'::uuid,
-  '{"tldr":"<one sentence>","assumptions":[],"deviations":[],"uncertainty":[],"detail":"<what was fixed>","flags":[]}'::jsonb
+  '{"tldr":"<one sentence>","assumptions":[],"deviations":[],"uncertainty":[],"detail":"<what was fixed>","flags":[],"transactional":false,"memory_records":[]}'::jsonb
 );
 
 SELECT update_work_item_status('<id>'::uuid, 'done',
@@ -101,6 +101,11 @@ SELECT update_work_item_status('<id>'::uuid, 'done',
 );
 ```
 
+If the fix touched a transactional flow, API boundary, or package contract, set `"transactional": true` and populate `memory_records` (see `agent-bootstrap.md`), then run:
+```bash
+node scripts/work-item-memory.mjs <work-item-id>
+```
+
 ```bash
 rm {USER_HOME}/.claude/fixit.marker
 ```