@lifeaitools/rdc-skills 0.35.4 → 0.35.5

package/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "rdc",
-  "version": "0.35.4",
+  "version": "0.35.5",
   "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/commands/build.md

@@ -69,6 +69,25 @@ Read the task title and description, then:
 - Mentions visualization, chart, diagram, SVG → `viz`
 - Multiple types? Dispatch multiple agents, each with its guide.
 
+### Execution primitive for create/open/build/deploy checklist rows
+
+When a dispatched agent's checklist row is to materialize a product shape,
+open a signed edit session, run a target's declared build gates, or deploy
+to dev-PM2/npm-registry, it uses the real, tested `rdc-harness` CLI instead
+of hand-rolled bash/curl:
+
+```bash
+node C:/Dev/rdc-harness/bin/rdc-harness.mjs <create|open|edit|build|deploy> <slug> --monorepo-root <the dispatched agent's own worktree>
+```
+
+One JSON receipt per call, exit 0/1 — tick the checklist row with the parsed
+receipt as evidence, not the raw dump. No Coolify awareness (production
+deploy stays `/rdc:deploy`'s own path) and no live co-editing surface
+outside `site-html`/`site-ts` (other classes get file-boundary save only —
+real, currently-unbuilt gap for other product classes, not something to
+paper over here). `open`/`edit` require `RDC_HARNESS_ISSUER_SECRET` set
+explicitly per-session — never a default.
+
 ## Procedure
 
 1. **Load the epic and its durable admission decisions:**

package/commands/deploy.md

@@ -27,11 +27,30 @@ No raw MCP dumps. No UUIDs unless asked.
 - `rdc:deploy audit` — fleet-wide scan for missed failures
 - `rdc:deploy audit --fix` — fleet scan + auto-remediate safe issues
 - `rdc:deploy maintenance <service>` — create, update, or verify a template-declared private service; no public domain or host port
+- `rdc:deploy dev <slug>` — deploy to PM2 dev (Vultr), not Coolify — see Mode 0
 - `rdc:deploy` (no args) — print mode menu, ask which
 
 ## Modes
 
-### Mode 1 — deploy <slug> [build-id]
+### Mode 0 — dev <slug> (PM2 development, not Coolify)
+
+PM2 dev deploys route through the real, tested `rdc-harness` CLI instead of
+raw PM2/curl — it already implements this path (`shipRoute: pm2-development`,
+via `@lifeaitools/regen-deploy-mgr` on loopback :52438, never raw PM2):
+
+```
+rdc:deploy dev: <slug>
+[ ] node C:/Dev/rdc-harness/bin/rdc-harness.mjs deploy <slug> --monorepo-root <caller's own worktree>
+[ ] JSON receipt parsed — shipRoute confirmed "pm2-development" (else: not this product's route, see receipt.reason)
+[ ] Receipt reports ok / the specific refusal, reported verbatim — not narrated
+✅ rdc:deploy dev: <slug> — <receipt outcome in one line>
+```
+
+A receipt with `applicable: false, reason: 'not_pm2_shipped'` means this slug
+ships a different way (registry, static, or Coolify) — report that plainly,
+do not retry as Coolify without confirming that's actually the right route.
+
+### Mode 1 — deploy <slug> [build-id] (Coolify — staging/production)
 
 ```
 rdc:deploy: <slug> → <domain>

package/commands/fixit.md

@@ -70,6 +70,23 @@ This signals the Stop hook that fixit is handling its own documentation.
 
 ### 4. Make the fix
 
+If the fix is a create/open/build/deploy step against a real fleet
+repository (materialize a product shape, open a signed edit session, run a
+target's declared build gates, or deploy to dev-PM2/npm-registry), use the
+real, tested `rdc-harness` CLI instead of hand-rolled bash/curl:
+
+```bash
+node C:/Dev/rdc-harness/bin/rdc-harness.mjs <create|open|edit|build|deploy> <slug> --monorepo-root <your own worktree, never the shared checkout>
+```
+
+One JSON receipt on stdout, exit 0/1 — use it as the evidence for whichever
+checklist/commit step it satisfies. It has no Coolify awareness (deploy here
+means PM2 dev or npm publish only — Coolify stays this skill's own §5.5 path)
+and no live co-editing surface outside `site-html`/`site-ts` targets (other
+classes get boundary-checked file save only). `open`/`edit` need
+`RDC_HARNESS_ISSUER_SECRET` set explicitly — no default exists or should.
+For anything not create/open/build/deploy shaped, edit files directly as below.
+
 Do the minimal work, conforming to `docs/CODING-STANDARDS.md` where present
 (regen-root; skip if absent). Scope creep rule: if you discover the fix requires more than originally scoped, **stop immediately**:
 1. Close the work item: `update_work_item_status('<id>', 'blocked', '["Escalated — scope exceeded fixit threshold"]')`

package/commands/flow.md

@@ -0,0 +1,94 @@
+---
+name: flow
+description: >-
+  Usage `rdc:flow [status|<state> ["<reason>"]|normal]` — show or set the current flow state. Flow is the single FSM surface for both work-shape (plan/design/collab/build/refactor/overnight/harness-testing) and ceremony relaxation (hotfix/maintenance). Supersedes `rdc:mode` — hotfix/maintenance are flow values now, not a separate system.
+---
+
+> **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`
+> Checklist-only output. No tool-call narration. No raw MCP/JSON/log dumps.
+> One checklist upfront, updated in place, shown again at end with a 1-line verdict.
+
+# rdc:flow — the one FSM surface
+
+## When to Use
+- Starting a turn — declare what kind of work this is (`build` needs a work item, `plan`/`design`/`collab` don't)
+- A production incident where process ceremony is costing minutes that matter (`hotfix`)
+- Routine upkeep that needs the same relaxation, without calling it an incident (`maintenance`)
+- Checking what's currently declared, and why
+
+## The one thing to understand first
+
+**There used to be two systems — mode (safety relaxation) and flow (work-shape declaration) — kept deliberately separate.** Operator instruction, 2026-08-25/26, direct and live: *"no special modes -- flow states only -- move hotfix to a flow state -- move env-bypass to a flow state -- the fsm will set what is enabled."*
+
+`hotfix` and `maintenance` are now real flow values, carrying the exact same safety guarantee they always did as modes: a reason is required to enter either, and only the named ceremony allowlist (`HOTFIX_DISABLES`/`HOTFIX_RELAXES_GATES` in `$LIFEAI_ENV/hooks/lib/rdc-flow.mjs` and `C:/Dev/rdc-harness/fsm-daemon/src/matrix.js`) is ever relaxed — never a safety guard (`execution-scope`, `rm-rf-danger`, `push-force`, credential-exposure, etc.).
+
+`rdc:mode` still exists for backward compatibility — the old mode axis is untouched, working code — but new work should use `rdc:flow` for everything, including what used to be a mode change.
+
+## States
+
+| State | Meaning | Requires a reason | Work item required |
+|---|---|---|---|
+| `plan` / `design` / `collab` | Conversational — nothing shipping | no | no |
+| `build` / `refactor` / `overnight` | Shipping code | no | **yes** |
+| `harness-testing` | Proving/testing rdc-harness's own binding guard | no | yes |
+| `hotfix` | Incident — ceremony relaxed for a bounded reason | **yes** | no |
+| `maintenance` | Routine upkeep — same relaxation power as hotfix, different label | **yes** | no |
+
+## Usage
+
+```
+rdc:flow                          # status — the default
+rdc:flow status
+rdc:flow build                    # declare shipping work — no reason needed
+rdc:flow hotfix "prod checkout 500s"   # relax ceremony — reason required
+rdc:flow normal                   # clear — back to no flow declared
+```
+
+## Steps
+
+### 1. Read the current state
+
+```bash
+node -e "import('file:///C:/Dev/lifeai-env/hooks/lib/rdc-flow.mjs').then(m=>console.log(JSON.stringify(m.currentState(),null,2)))"
+```
+
+`currentState()` reads BOTH axes at once — the declared flow AND any live `rdc-mode.mjs` hotfix/maintenance window — because a relaxed mode overrides a required flow (loosest state wins) and a status check that only showed flow would miss that override entirely. Report `flow` (or `null` if nothing declared — the fail-closed default), `flowReason`/`flowSetBy`/`flowSetAt` when set, and `mode` (`normal` unless a hotfix/maintenance window is open), with `modeReason`/`modeMinutesLeft` when it isn't.
+
+### 2. Setting a plain work-shape flow — no reason required
+
+```bash
+node -e "import('file:///C:/Dev/lifeai-env/hooks/lib/rdc-flow.mjs').then(m=>console.log(JSON.stringify(m.setFlow(process.argv[1],{setBy:process.argv[2]}),null,2)))" "<flow>" "<session-id>"
+```
+
+### 3. Setting hotfix or maintenance — a reason is REQUIRED
+
+Refuse to proceed without one — `setFlow` throws on an empty reason for these two values, by design, same as `setMode` always did: an unexplained disarm is how a temporary state becomes permanent.
+
+```bash
+node -e "import('file:///C:/Dev/lifeai-env/hooks/lib/rdc-flow.mjs').then(m=>console.log(JSON.stringify(m.setFlow('hotfix',{reason:process.argv[1],setBy:process.argv[2]}),null,2)))" "<reason>" "<session-id>"
+```
+
+Then state plainly, in the checklist: **which guards are now relaxed, that the safety set is still armed, and that this stays active until explicitly cleared (no TTL — active management, not a silent timer).**
+
+### 4. Returning to normal
+
+```bash
+node -e "import('file:///C:/Dev/lifeai-env/hooks/lib/rdc-flow.mjs').then(m=>{m.clearFlow();console.log('cleared')})"
+```
+
+Do this **as soon as the incident is over**, or the moment the declared work-shape changes.
+
+## Rules
+
+- **A reason is mandatory for `hotfix`/`maintenance`.** Every other flow value is a plain, reason-free declaration.
+- **No TTL, ever, by default.** A flow — including `hotfix`/`maintenance` — stays active until explicitly cleared or replaced. Never a silent expiry.
+- **Fail-closed.** No flow declared, an unreadable daemon, or a malformed response all read as `{flow: null}` — same behavior a work-shipping turn always had before this mechanism existed.
+- **Never widen the relaxation set to unblock yourself.** A guard blocking a hotfix flow is information. Editing the allowlist to get past a red guard is the same violation as editing code to satisfy a failing gate (`.claude/rules/debugging-protocol.md` Rule 10).
+- **hotfix/maintenance flow is not a bypass of review.** Work done under it still needs its work item, its evidence, and its close.
+
+## Verification
+
+```bash
+node --test $LIFEAI_ENV/tests/rdc-flow.test.mjs
+node --test C:/Dev/rdc-harness/fsm-daemon/tests/matrix.test.js
+```

package/commands/mode.md

@@ -8,6 +8,12 @@ description: >-
 > Checklist-only output. No tool-call narration. No raw MCP/JSON/log dumps.
 > One checklist upfront, updated in place, shown again at end with a 1-line verdict.
 
+> **Superseded by `rdc:flow`, 2026-08-26.** `hotfix`/`maintenance` are now
+> flow values (same reason-required, same relaxation allowlist, no TTL by
+> default) — one FSM surface instead of two. This skill's underlying code
+> (`rdc-mode.mjs`) is untouched and still works; new work should use
+> `rdc:flow hotfix "<reason>"` instead of `rdc:mode hotfix ...`.
+
 # rdc:mode — operating mode
 
 ## When to Use

package/commands/release.md

@@ -39,6 +39,17 @@ rdc:release: <repo> vX.Y.Z -> vA.B.C
 - Never force push or bypass hooks.
 - Never declare success without verifying the installed or deployed version.
 
+For a `package`-class target that already resolves through `rdc-harness`
+(a real monorepo subtree, not a standalone repo like this one), its
+`packages/deploy/src/runners/registry-release.mjs` runner already proves the
+"Tests/self-test passed" through "Local install/update executed" steps
+safely — real `npm pack`, isolated-prefix install (never the real global
+store), real verify, and `--live` explicitly gates the actual publish. Where
+applicable, `node C:/Dev/rdc-harness/bin/rdc-harness.mjs deploy <slug>
+[--live]` can supply those checklist rows' evidence directly instead of
+hand-rolling the same pack/install/verify cycle. This does not replace
+version bump/tag/push — the harness CLI does neither.
+
 ## RDC Skills Package
 
 After publishing this package to npm, a clean-box install should use:

package/guides/agent-bootstrap.md

@@ -1,11 +1,11 @@
-# Agent Bootstrap — Read This First
-> Every dispatched agent reads this before their role-specific guide.
-> Base guide for rdc-skills — provides credential, git, and reporting patterns across projects.
-
----
-
-## Who You Are
-
+# Agent Bootstrap — Read This First
+> Every dispatched agent reads this before their role-specific guide.
+> Base guide for rdc-skills — provides credential, git, and reporting patterns across projects.
+
+---
+
+## Who You Are
+
 You are a subagent dispatched by the rdc:build supervisor. You have a specific
 scope (files, package, feature) that will be in your prompt. Stay in that scope.
 NEVER modify files outside it.
@@ -15,198 +15,242 @@ RDC implementation posture for assumptions, minimal changes, surgical scope,
 verification evidence, and escalation.
 
 ---
-
-## Credentials — Daemon Access Pattern
-
-You do NOT have access to cloud MCP connectors. Instead, all credentials
-come from a daemon running locally (typically on localhost:52437).
-
-**Ping first to confirm availability:**
-```bash
-curl -s http://127.0.0.1:52437/ping
-```
-If it doesn't respond — report BLOCKED, do not proceed.
-
-**Get a credential:**
-```bash
-curl -s http://127.0.0.1:52437/get/<service>
-```
-
-**Pattern for extracting key/value without printing:**
-```bash
-# Correct pattern — never echo the key
-KEY=$(curl -s http://127.0.0.1:52437/get/<service> | python3 -c "import sys,json; print(json.load(sys.stdin)['key'])")
-curl -s -H "Authorization: Bearer $KEY" https://api.example.com/...
-```
-
-**Never print credentials to stdout.** Capture to a variable, use inline, discard.
-
----
-
-## Project Directory Convention
-
-This plugin uses the `.rdc/` directory convention. Check for it first:
-
-```bash
-# Check if .rdc/ exists at project root
-ls {PROJECT_ROOT}/.rdc/config.json 2>/dev/null && echo "using .rdc/" || echo "using docs/ fallback"
-```
-
-**Path resolution rule:**
-- Guides: `{PROJECT_ROOT}/.rdc/guides/` → fallback: `{PROJECT_ROOT}/docs/guides/`
-- Plans: `{PROJECT_ROOT}/.rdc/plans/` → fallback: `{PROJECT_ROOT}/docs/plans/`
-- Reports: `{PROJECT_ROOT}/.rdc/reports/` → fallback: `{PROJECT_ROOT}/docs/reports/`
-- Research: `{PROJECT_ROOT}/.rdc/research/` → fallback: `{PROJECT_ROOT}/docs/research/`
-
-If `.rdc/config.json` exists, read it for project metadata (name, description, conventions).
-
----
-
-## Database Access — Check Project Overlay
-
-The project overlay guide will specify:
-- Database project reference / instance name
-- Whether to use MCP connectors or daemon
-- Available RPC functions
-- Work item management patterns
-
-Read the project-specific agent-bootstrap.md overlay for exact connection details.
-
----
-
-## Git Rules
-
-- Branch: Always use the project's primary development branch (typically `develop` or `main`)
-- Auto-commit after completing your scope — no confirmation needed
-- Commit message must use conventional format: `feat/fix/chore/refactor(<scope>): description`
-- Push to origin after committing
-- NEVER force-push
-
----
-
+
+## Credentials — Daemon Access Pattern
+
+You do NOT have access to cloud MCP connectors. Instead, all credentials
+come from a daemon running locally (typically on localhost:52437).
+
+**Ping first to confirm availability:**
+```bash
+curl -s http://127.0.0.1:52437/ping
+```
+If it doesn't respond — report BLOCKED, do not proceed.
+
+**Get a credential:**
+```bash
+curl -s http://127.0.0.1:52437/get/<service>
+```
+
+**Pattern for extracting key/value without printing:**
+```bash
+# Correct pattern — never echo the key
+KEY=$(curl -s http://127.0.0.1:52437/get/<service> | python3 -c "import sys,json; print(json.load(sys.stdin)['key'])")
+curl -s -H "Authorization: Bearer $KEY" https://api.example.com/...
+```
+
+**Never print credentials to stdout.** Capture to a variable, use inline, discard.
+
+---
+
+## Project Directory Convention
+
+This plugin uses the `.rdc/` directory convention. Check for it first:
+
+```bash
+# Check if .rdc/ exists at project root
+ls {PROJECT_ROOT}/.rdc/config.json 2>/dev/null && echo "using .rdc/" || echo "using docs/ fallback"
+```
+
+**Path resolution rule:**
+- Guides: `{PROJECT_ROOT}/.rdc/guides/` → fallback: `{PROJECT_ROOT}/docs/guides/`
+- Plans: `{PROJECT_ROOT}/.rdc/plans/` → fallback: `{PROJECT_ROOT}/docs/plans/`
+- Reports: `{PROJECT_ROOT}/.rdc/reports/` → fallback: `{PROJECT_ROOT}/docs/reports/`
+- Research: `{PROJECT_ROOT}/.rdc/research/` → fallback: `{PROJECT_ROOT}/docs/research/`
+
+If `.rdc/config.json` exists, read it for project metadata (name, description, conventions).
+
+---
+
+## Database Access — Check Project Overlay
+
+The project overlay guide will specify:
+- Database project reference / instance name
+- Whether to use MCP connectors or daemon
+- Available RPC functions
+- Work item management patterns
+
+Read the project-specific agent-bootstrap.md overlay for exact connection details.
+
+---
+
+## Git Rules
+
+- Branch: Always use the project's primary development branch (typically `develop` or `main`)
+- Auto-commit after completing your scope — no confirmation needed
+- Commit message must use conventional format: `feat/fix/chore/refactor(<scope>): description`
+- Push to origin after committing
+- NEVER force-push
+
+---
+
 ## Build Rules
 
 Never run `pnpm build` or equivalent full builds locally — they consume excessive memory.
 Type-check only: `npx tsc --noEmit --project <path>/tsconfig.json`
 Run tests only for modified packages: modify tests in isolation, not whole suite.
 
-### No Foreground Windows
+### Terminal/process launches — caller-logged, not hard-blocked
 
-Agent-launched processes must not steal focus. This is a hard local-operator
-rule, not a preference.
+**Narrowed 2026-08-26** (epic 688ad6da, lifeai-env; direct operator
+instruction: "remove the PreToolUse foreground-window guard entirely...
+replace it with caller-logging for traceability"). The old hard block on
+every raw `Start-Process`/`cmd /c start`/window-focus API/bare `.ps1` launch
+is retired — `foreground-process-gate.js` (rdc-skills) no longer enforces any
+of it. The replacement isn't a weaker rule; it's a different mechanism:
+lifeai-env's `lib/TermLaunch.psm1` (`Invoke-TermLaunchHidden` /
+`Invoke-TermLaunchInteractive`) logs every caller (script + line + argv) to
+`C:/Dev/.logs` **before** spawning, for every launch — which answers "who
+launched this and with what" for every case, not just the ones a regex
+pattern happened to catch.
 
-- Playwright must run headless. Do not use `--headed`, `--ui`, `codegen`, `open`, `show-report`, or `PWDEBUG=1` in agent sessions.
-- Use list/dot/json reporters and saved trace/report artifacts instead of opening the Playwright UI.
-- PowerShell helpers must use `-WindowStyle Hidden -NonInteractive`, or a hidden wrapper.
-- `Start-Process` must include `-WindowStyle Hidden` or `-WindowStyle Minimized`.
-- `cmd /c start` must use `/min` for intentionally visible tools or `/b` for background work.
-- Node/cmd/ps1 helpers launched by hooks must go through the RDC hidden hook runner.
+- Use `Invoke-TermLaunchHidden -Command <cmd> [-Arguments][-WorkingDirectory]`
+  for a background/no-window process, and `Invoke-TermLaunchInteractive
+  [-Title]` for a visible new terminal tab — never a raw `Start-Process`/
+  `wt.exe` call. Neither primitive accepts an arbitrary inline multi-line
+  command string, only a flat command + argument array or a `-File <script>`
+  path — this structurally prevents the `wt.exe -Command` argv-mangling bug
+  class, which no amount of pattern-blocking ever fully closed.
+- Node/cmd/ps1 helpers launched by hooks must still go through the RDC hidden
+  hook runner (`hooks/run-bash-hidden.ps1` and its RdcRun contract) — that
+  convention is unchanged.
+
+**One safety property survives, unrelated to the above and never retired:**
+Playwright must still run headless in agent sessions. Do not use `--headed`,
+`--ui`, `codegen`, `open`, `show-report`, or `PWDEBUG=1`. Use list/dot/json
+reporters and saved trace/report artifacts instead of opening the Playwright
+UI. `foreground-process-gate.js` still hard-blocks this — it was never a
+terminal-launch-primitive question, so narrowing the launch rules above never
+touched it.
 
 Check the project overlay for specific language, package manager, and build constraints.
-
----
-
-## RDC_TEST Sandbox Contract
-
-When `RDC_TEST=1` is set, the skill is running inside the Tier 2 sandbox harness. Short-circuit all destructive external side effects.
-
-**How to detect:** `process.env.RDC_TEST === '1'` (Node/JS) or `[ "$RDC_TEST" = "1" ]` (bash). The Tier 2 runner sets this before invoking `claude --print`.
-
-**Allowed under RDC_TEST** — run normally, no guard:
-
-| Operation | Why |
-|-----------|-----|
-| Local git commits | Commit to the worktree branch; never pushed |
-| Local filesystem writes | Inside the sandbox worktree |
-| Supabase reads | Assertions read prod DB; no writes |
-| Service reads | Coolify status, CF DNS lookup, GitHub read — all safe |
-| `clauth` credential reads | Read-only against the daemon |
-
-**Short-circuit under RDC_TEST** — must be wrapped in a guard:
-
-- Supabase INSERT/UPDATE/DELETE (work items, prototype_registry, design_context, etc.) — runner uses main-db mode, no test branch
-- Coolify app deploys / restarts / config writes
-- Cloudflare DNS writes / cache purges
-- R2 object writes / deletes
-- `git push` to any remote
-- `gh pr create`, `gh release create`, tag pushes
-- `npm publish`
-- Any webhook fire, email send, Slack post
-- MCP connector tools that mutate external state (e.g. `mcp__coolify__deploy`)
-
-**Bash pattern:**
-```bash
-if [ "$RDC_TEST" != "1" ]; then
+
+### Harness Use — Global Policy (applies to every rdc:* skill)
+
+If ANY step of the skill you are executing — regardless of which skill —
+needs to materialize a real product shape, open a signed edit session, run a
+target's own declared build gates, or deploy to dev-PM2/npm-registry, you
+MUST use the real, tested `rdc-harness` CLI instead of hand-rolled bash/curl:
+
+```bash
+node C:/Dev/rdc-harness/bin/rdc-harness.mjs <create|open|edit|build|deploy> <slug> --monorepo-root <your own worktree, never the shared checkout>
+```
+
+One JSON receipt per call, exit 0/1 — use it as evidence for whatever
+checklist/commit step it satisfies. This is a genuinely FORCED requirement,
+not a suggestion: a skill that hand-rolls its own create/open/build/deploy
+logic when this applies is doing exactly the duplicated, unproven work this
+policy exists to stop.
+
+This does NOT apply to a skill with no create/open/build/deploy step at all
+(status/report/help/analysis/media/conversation skills) — do not invent one.
+`open`/`edit` require `RDC_HARNESS_ISSUER_SECRET` set explicitly, no default.
+`deploy` has zero Coolify awareness (Coolify stays `rdc:deploy`'s own path)
+and no live co-editing surface outside `site-html`/`site-ts` targets.
+
+---
+
+## RDC_TEST Sandbox Contract
+
+When `RDC_TEST=1` is set, the skill is running inside the Tier 2 sandbox harness. Short-circuit all destructive external side effects.
+
+**How to detect:** `process.env.RDC_TEST === '1'` (Node/JS) or `[ "$RDC_TEST" = "1" ]` (bash). The Tier 2 runner sets this before invoking `claude --print`.
+
+**Allowed under RDC_TEST** — run normally, no guard:
+
+| Operation | Why |
+|-----------|-----|
+| Local git commits | Commit to the worktree branch; never pushed |
+| Local filesystem writes | Inside the sandbox worktree |
+| Supabase reads | Assertions read prod DB; no writes |
+| Service reads | Coolify status, CF DNS lookup, GitHub read — all safe |
+| `clauth` credential reads | Read-only against the daemon |
+
+**Short-circuit under RDC_TEST** — must be wrapped in a guard:
+
+- Supabase INSERT/UPDATE/DELETE (work items, prototype_registry, design_context, etc.) — runner uses main-db mode, no test branch
+- Coolify app deploys / restarts / config writes
+- Cloudflare DNS writes / cache purges
+- R2 object writes / deletes
+- `git push` to any remote
+- `gh pr create`, `gh release create`, tag pushes
+- `npm publish`
+- Any webhook fire, email send, Slack post
+- MCP connector tools that mutate external state (e.g. `mcp__coolify__deploy`)
+
+**Bash pattern:**
+```bash
+if [ "$RDC_TEST" != "1" ]; then
   curl -X POST "$DEPLOY_API_BASE/api/v1/applications/$UUID/deploy" ...
-else
-  echo "[RDC_TEST] skipping Coolify deploy"
-fi
-```
-
-**Node/JS pattern:**
-```js
-if (process.env.RDC_TEST !== '1') {
-  await deployToCoolify(...);
-} else {
-  console.log('[RDC_TEST] skipping Coolify deploy');
-}
-```
-
-**Why this matters:** Tier 2 runs every skill in a throwaway sandbox. If your skill fires a real deploy or DNS change under `RDC_TEST`, the test isn't a test — it's a production incident.
-
-**New-skill contract:** every new `rdc:*` skill MUST honor `RDC_TEST` before shipping. Tier 2 manifests will fail any skill that writes to external state under the flag.
-
+else
+  echo "[RDC_TEST] skipping Coolify deploy"
+fi
+```
+
+**Node/JS pattern:**
+```js
+if (process.env.RDC_TEST !== '1') {
+  await deployToCoolify(...);
+} else {
+  console.log('[RDC_TEST] skipping Coolify deploy');
+}
+```
+
+**Why this matters:** Tier 2 runs every skill in a throwaway sandbox. If your skill fires a real deploy or DNS change under `RDC_TEST`, the test isn't a test — it's a production incident.
+
+**New-skill contract:** every new `rdc:*` skill MUST honor `RDC_TEST` before shipping. Tier 2 manifests will fail any skill that writes to external state under the flag.
+
 **Known blocker:** Project-specific cwd hooks must check `process.env.RDC_TEST === '1'` and call `process.exit(0)` early to allow Tier 2 sandbox runs. Without this bypass, headless self-test invocations can fail before the skill loads. File: `~/.claude/hooks/check-cwd.js`.
-
----
-
-## Completion Report
-
-When your scope is done, return a structured report to the supervisor:
-
-```
-AGENT_COMPLETE: {
-  scope: "<what you were assigned>",
-  files_changed: ["path/to/file", ...],
-  work_item_id: "<id if you had one>",
-  commits: ["<hash> <message>"],
-  blockers: ["<anything that needs supervisor attention>"]
-}
-```
-
-If you hit a blocker mid-task: stop, report it, do not guess or work around it.
-
----
-
-## Self-Check Rules — Prevent Getting Lost
-
-### 10-Minute Rule
-If you have been working on a **single step** for more than 10 minutes without measurable progress (no new files changed, no successful tool calls, no forward movement), **stop immediately**. Do not keep trying variations. Report it as a blocker.
-
-### 2-Retry Rule
-If the **same command or approach fails twice**, stop. Do not attempt a third variation or creative workaround. Report the failure with the exact error output.
-
-### Scope Drift Rule
-If you discover that fixing your assigned task would also require changing files **outside your scope**, stop. Do not fix them. Add them to `blockers` in your AGENT_COMPLETE report. The supervisor assigns them separately.
-
-### What "measurable progress" means
-- A file was created or modified ✅
-- A tool call succeeded and returned useful data ✅
-- A command ran without error ✅
-- Trying the same thing with slightly different parameters ❌
-- Reading the same file again hoping for different insight ❌
-- Rephrasing a failing query ❌
-
----
-
+
+---
+
+## Completion Report
+
+When your scope is done, return a structured report to the supervisor:
+
+```
+AGENT_COMPLETE: {
+  scope: "<what you were assigned>",
+  files_changed: ["path/to/file", ...],
+  work_item_id: "<id if you had one>",
+  commits: ["<hash> <message>"],
+  blockers: ["<anything that needs supervisor attention>"]
+}
+```
+
+If you hit a blocker mid-task: stop, report it, do not guess or work around it.
+
+---
+
+## Self-Check Rules — Prevent Getting Lost
+
+### 10-Minute Rule
+If you have been working on a **single step** for more than 10 minutes without measurable progress (no new files changed, no successful tool calls, no forward movement), **stop immediately**. Do not keep trying variations. Report it as a blocker.
+
+### 2-Retry Rule
+If the **same command or approach fails twice**, stop. Do not attempt a third variation or creative workaround. Report the failure with the exact error output.
+
+### Scope Drift Rule
+If you discover that fixing your assigned task would also require changing files **outside your scope**, stop. Do not fix them. Add them to `blockers` in your AGENT_COMPLETE report. The supervisor assigns them separately.
+
+### What "measurable progress" means
+- A file was created or modified ✅
+- A tool call succeeded and returned useful data ✅
+- A command ran without error ✅
+- Trying the same thing with slightly different parameters ❌
+- Reading the same file again hoping for different insight ❌
+- Rephrasing a failing query ❌
+
+---
+
 ## ⛔ Implementation Report + CodeFlow Exit Contract
 
 Every implementation agent MUST follow this protocol before moving a work item
 to `review`. Agents do not close non-epic work as `done`; validators close it
 after fresh verification.
-
-### Step 1 — Tick checklist items as you complete them
-
+
+### Step 1 — Tick checklist items as you complete them
+
 ```sql
 SELECT update_checklist_item(
   '<work-item-id>'::uuid,
@@ -221,10 +265,10 @@ SELECT update_checklist_item(
 Call this for each item AS you complete it — not all at once at the end. The
 database records every tick in `work_item_checklist_events`. Supervisor and
 validator re-ticks are rejected by the exit gate.
-
-### Step 2 — Submit implementation report BEFORE marking done
-
-```sql
+
+### Step 2 — Submit implementation report BEFORE marking done
+
+```sql
 SELECT submit_implementation_report(
   '<work-item-id>'::uuid,
   '{
@@ -276,20 +320,20 @@ SELECT update_work_item_status(
 If any `required: true` checklist item is still unchecked, was re-ticked by a
 supervisor/validator, or was ticked by a different session than the originating
 agent, the DB and PreToolUse hook reject the close.
-
-### Supervisor workflow
-
-- All zeros → clean run, proceed
-- `flags_count > 0` or `deviations_count > 0` → pull full report:
-  ```sql
-  SELECT implementation_report FROM work_items WHERE id = '<id>';
-  ```
-
----
-
-## Now read your role-specific guide
-
-Path: `{PROJECT_ROOT}/.rdc/guides/<type>.md` (e.g., `frontend.md`, `backend.md`, `data.md`)
-Fallback: `{PROJECT_ROOT}/docs/guides/<type>.md` if `.rdc/` does not exist.
-
-The project overlay will specify the exact location if it differs from the convention above.
+
+### Supervisor workflow
+
+- All zeros → clean run, proceed
+- `flags_count > 0` or `deviations_count > 0` → pull full report:
+  ```sql
+  SELECT implementation_report FROM work_items WHERE id = '<id>';
+  ```
+
+---
+
+## Now read your role-specific guide
+
+Path: `{PROJECT_ROOT}/.rdc/guides/<type>.md` (e.g., `frontend.md`, `backend.md`, `data.md`)
+Fallback: `{PROJECT_ROOT}/docs/guides/<type>.md` if `.rdc/` does not exist.
+
+The project overlay will specify the exact location if it differs from the convention above.

package/hooks/foreground-process-gate.js

@@ -1,9 +1,28 @@
 #!/usr/bin/env node
 /**
- * PreToolUse hook — block focus-stealing foreground process launches.
+ * PreToolUse hook — blocks Playwright headed/UI-mode invocations only.
+ *
+ * NARROWED 2026-08-26 (epic 688ad6da WP-4, lifeai-env). This file used to
+ * ALSO block every raw foreground process launch (Start-Process without
+ * -WindowStyle Hidden, cmd /c start without /min, window-focus Win32 APIs,
+ * bare PowerShell .ps1 launches) -- that entire class is retired per Dave's
+ * direct operator instruction: "remove the PreToolUse foreground-window
+ * guard entirely... replace it with caller-logging for traceability."
+ * lifeai-env's lib/TermLaunch.psm1 (Invoke-TermLaunchHidden /
+ * Invoke-TermLaunchInteractive) is the replacement -- it logs every caller
+ * (script + line + argv) to C:/Dev/.logs BEFORE spawning, which is strictly
+ * more informative than a hard block that told a caller only "add
+ * -WindowStyle Hidden" and never recorded who asked.
+ *
+ * checkPlaywright is a SEPARATE, unrelated concern (Design Decision D3,
+ * .rdc/plans/terminal-launch-consolidation.md in lifeai-env): agent
+ * sessions must never pop an interactive Playwright UI. That has nothing to
+ * do with terminal/process launch primitives, so it was deliberately kept
+ * here rather than folded into caller-logging, which would have silently
+ * dropped a real safety property this narrowing was never asked to remove.
+ * The filename is legacy -- kept to avoid an unrelated hookify-manifest
+ * rewire for a rename that changes nothing about what the file does.
  */
-'use strict';
-
 const hookLog = require('./hook-logger');
 
 function readStdin() {
@@ -35,35 +54,6 @@ function toolText(raw) {
   try { return JSON.stringify(raw.tool_input || raw); } catch { return ''; }
 }
 
-function hasHiddenIntent(command) {
-  return /-WindowStyle\s+Hidden/i.test(command) ||
-    /-WindowStyle\s+Minimized/i.test(command) ||
-    /windowsHide\s*:\s*true/i.test(command) ||
-    /CreateNoWindow\s*=\s*\$?true/i.test(command) ||
-    /Start-Job\b/i.test(command) ||
-    /--background\b/i.test(command) ||
-    /\bHEADLESS\s*=\s*(1|true)\b/i.test(command) ||
-    /\bCI\s*=\s*(1|true)\b/i.test(command);
-}
-
-function hasExplicitWindowOverride(command) {
-  return /\bRDC_ALLOW_WINDOW_FOCUS\s*=\s*(1|true)\b/i.test(command) ||
-    /\bRDC_INTERACTIVE_WINDOW\s*=\s*(1|true)\b/i.test(command);
-}
-
-function checkWindowFocusApi(command) {
-  if (hasExplicitWindowOverride(command)) return;
-  const focusApi = /\b(SetForegroundWindow|SwitchToThisWindow|AppActivate|SetWindowPos|ShowWindowAsync?|BringWindowToTop)\b/i;
-  const broadWindowApi = /\b(EnumWindows|Get-Process\s+\|\s*Where-Object|GetWindow|FindWindow)\b/i;
-  const windowMutation = /\b(minimi[sz]e|restore|foreground|focus|activate|collapse)\b/i;
-  if (focusApi.test(command) || (broadWindowApi.test(command) && windowMutation.test(command))) {
-    block(
-      'Window focus/restore/minimize/collapse operations are not allowed in agent-launched commands. Spawn helpers hidden/no-window instead; set RDC_ALLOW_WINDOW_FOCUS=1 only for an explicitly requested interactive recovery action.',
-      { kind: 'window-focus-api' },
-    );
-  }
-}
-
 function checkPlaywright(command) {
   if (!/\b(playwright|@playwright\/test)\b/i.test(command)) return;
 
@@ -82,45 +72,13 @@ function checkPlaywright(command) {
   }
 }
 
-function checkPowerShell(command) {
-  if (!/\bStart-Process\b/i.test(command)) return;
-  if (hasHiddenIntent(command)) return;
-  block(
-    '`Start-Process` must include `-WindowStyle Hidden` or `-WindowStyle Minimized` for agent-launched node/cmd/ps1/test processes. Focus/restore/collapse APIs remain blocked unless explicitly requested.',
-    { kind: 'start-process' },
-  );
-}
-
-function checkCmdStart(command) {
-  if (!/\bcmd(?:\.exe)?\s+\/c\s+start\b/i.test(command)) return;
-  if (/\bcmd(?:\.exe)?\s+\/c\s+start\s+(""|''|`"")?\s*\/b\b/i.test(command)) return;
-  block(
-    '`cmd /c start` must use `/min` or `/b` for background tools. Focus/restore/collapse APIs remain blocked unless explicitly requested.',
-    { kind: 'cmd-start' },
-  );
-}
-
-function checkDirectShellLaunch(command) {
-  if (hasHiddenIntent(command)) return;
-  if (/\bpowershell(?:\.exe)?\b[^|\n]*(?:-File\s+[^|\n]*\.ps1|\.ps1\b)/i.test(command)) {
-    block(
-      'PowerShell script launches from agent tooling must use `-WindowStyle Hidden -NonInteractive` or a hidden wrapper.',
-      { kind: 'powershell-ps1' },
-    );
-  }
-}
-
 async function main() {
   let raw;
   try { raw = JSON.parse(await readStdin()); } catch { process.exit(0); }
   const command = toolText(raw);
   if (!command) pass({ reason: 'no-command' });
 
-  checkWindowFocusApi(command);
   checkPlaywright(command);
-  checkPowerShell(command);
-  checkCmdStart(command);
-  checkDirectShellLaunch(command);
 
   pass({ reason: 'clean' });
 }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@lifeaitools/rdc-skills",
-  "version": "0.35.4",
+  "version": "0.35.5",
   "description": "RDC typed-agent dispatch skill suite for Claude Code - plan, build, review, overnight builds",
   "keywords": [
     "claude-code",
@@ -44,7 +44,7 @@
     "validate": "node tests/validate-skills.js",
     "rdc-design": "node scripts/rdc-design-cli.mjs",
     "test:hooks": "node scripts/test-rdc-hooks.mjs",
-    "test:truth-gate": "node tests/run-evidence-gate.test.mjs && node tests/work-item-exit-gate-l2.test.mjs && node tests/work-item-exit-gate-l3.test.mjs && node tests/require-work-item-on-commit.test.mjs && node tests/harness-gates.test.mjs",
+    "test:truth-gate": "node tests/run-evidence-gate.test.mjs && node tests/work-item-exit-gate-l2.test.mjs && node tests/work-item-exit-gate-l3.test.mjs && node tests/require-work-item-on-commit.test.mjs && node tests/harness-gates.test.mjs && node tests/foreground-process-gate.test.mjs",
     "test:acceptance": "node tests/acceptance.test.mjs && node tests/install-rdc-skills.test.mjs && node tests/help-surface.test.mjs && node tests/manifest-contract-fields.test.mjs && node tests/plugin-namespace-names.test.mjs && node tests/skill-test-matrix.test.mjs && node tests/completion-gate-supervisor-admission.test.mjs && node tests/curl-surface.test.mjs && node tests/clauth-plugin-postinstall.test.mjs",
     "acceptance": "node scripts/acceptance.mjs --changed",
     "test:mcp": "node tests/mcp.test.mjs",

package/skills/architecture-reviewer/SKILL.md

@@ -13,6 +13,7 @@ description: >-
   rdc:review step 8b+, or standalone before merging a new package/module.
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 > **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`
 > Checklist-only output. No tool-call narration. No raw MCP/JSON/log dumps.
 > One checklist upfront, updated in place, shown again at end with a 1-line verdict.

package/skills/behavior-audit/SKILL.md

@@ -3,6 +3,7 @@ name: behavior-audit
 description: "Usage `rdc:behavior-audit <report-dir> [--since-days N] [--latest N] [--reprocess]` — produces a bounded, redacted Claude/Codex transcript evidence bundle, incrementally skips completed transcript hashes, and aligns candidate behavior problems to shared truth-governance rules."
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 > **OUTPUT CONTRACT:** Begin and end with the same checklist. Do not call an audit clean, complete, or compliant without the evidence-bundle manifest and an independent validator decision.
 
 # rdc:behavior-audit — Cross-Engine Truth and Behavior Audit

package/skills/brochure/SKILL.md

@@ -3,6 +3,7 @@ name: brochure
 description: "Usage `rdc:brochure <input> [--out <path>] [--template <name>] [--format Letter|A4]` — Turn a zip, folder, HTML file, URL, or markdown folder into a print-quality PDF brochure via Puppeteer. Auto-detects print-variant HTML, honors @page CSS, falls back to a Studio-token-aware template when no HTML exists."
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 > **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`
 > Checklist-only output. No tool-call narration. No raw MCP/JSON/log dumps.
 > One checklist upfront, updated in place, shown again at end with a 1-line verdict.

package/skills/brochurify/SKILL.md

@@ -12,6 +12,7 @@ triggers:
   - monkey_dispatch payload with skill="brochurify"
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 # rdc:brochurify Orchestrator
 > **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`
 > Report brochure job state, artifacts, and blockers directly; do not dump raw tool logs.

package/skills/clean-code-analyzer/SKILL.md

@@ -8,6 +8,7 @@ description: >-
   mechanical AST checks, not LLM judgment — see `scripts/clean-code-score.mjs`.
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 > **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`
 > Checklist-only output. No tool-call narration. No raw MCP/JSON/log dumps.
 > One checklist upfront, updated in place, shown again at end with a 1-line verdict.

package/skills/convert/SKILL.md

@@ -3,6 +3,7 @@ name: convert
 description: "Usage `rdc:convert <input> [--out <dir>] [--to markdown|word] [--images assets|base64|s3]` — Convert .docx/.pptx/.ppt → Markdown (Word OMML equations as KaTeX TeX, tables, images) or Markdown → Word via the build-corpus CLI from npm `regen.mde`. Portable: runs in any session that can reach npm — Claude Code CLI, Codex, and claude.ai can fetch + run it. Use whenever the user asks to convert an Office document, build a Markdown corpus from .docx/.pptx, turn Markdown into a .docx, or open the report in the regen-mde editor on Windows."
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 # rdc:convert — Office ↔ Markdown conversion (build-corpus) + regen-mde editor
 
 > **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`

package/skills/edit/SKILL.md

@@ -14,6 +14,18 @@ description: "Usage `rdc:edit <site|brand|route|file>` — open the local websit
 
 # rdc:edit — Local Website Editor Launcher
 
+**This skill is the designated handler `rdc-harness` dispatches to.**
+`packages/work/src/editors.mjs`'s `BY_CLASS` table maps `site-html`/`site-ts`
+product classes to editor id `'rdc:edit'` — when an agent calls
+`node C:/Dev/rdc-harness/bin/rdc-harness.mjs edit <slug>` against a website
+target, its receipt names `editorId: 'rdc:edit'` and an `editableBoundary`.
+This skill IS that handler: resolve the same target, launch/reuse the editor
+host, use the harness's `editableBoundary` as the save boundary if the call
+originated from a harness `edit` receipt. Every other product class
+(`app`/`package`/`mcp`/`model`/`artifact`) maps to editor id `'source'` — a
+boundary-checked file save with no live/co-editing surface at all; that gap
+is real and unbuilt, not something this skill covers.
+
 ## When to Use
 - The user wants to open a site, brand, route, or file in the local editor app
 - The user says "open this in the editor" or asks for the editor-host workflow

package/skills/env/SKILL.md

@@ -3,6 +3,7 @@ name: env
 description: "Usage `rdc:env [status|install|repair|update]` — Manage the LIFEAI environment harness: check status, install on a fresh box, repair broken services, or pull updates. Reads $LIFEAI_ENV/manifest.json as the source of truth. Use when: 'check the environment', 'install environment', 'repair environment', 'update environment', 'setup env', 'fix env', 'env status', or after a reboot/GPU crash."
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 > **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`
 > Checklist-only output. No tool-call narration. No raw MCP/JSON/log dumps.
 > One checklist upfront, updated in place, shown again at end with a 1-line verdict.

package/skills/extract-verifier-rules/SKILL.md

@@ -12,6 +12,7 @@ triggers:
   - nightly cron at 3:00 AM PT
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 # rdc:extract-verifier-rules
 > **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`
 > Return candidate rules, evidence, and PR status directly; do not dump raw tool logs.

package/skills/fs-mcp/SKILL.md

@@ -3,6 +3,7 @@ name: fs-mcp
 description: "Usage `rdc:fs-mcp <task>` — Use the File System MCP bridge for live repo reads, safe writes, cloud-to-local ingest, and GitHub-branch imports into a dirty local monorepo. Use when Claude.ai, Cowork, or CLI agents need fs_read/fs_write/fs_import_git_files guidance."
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 > **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`
 > Checklist-only output. No tool-call narration. No raw MCP/JSON/log dumps.
 > One checklist upfront, updated in place, shown again at end with a 1-line verdict.

package/skills/help/SKILL.md

@@ -3,6 +3,7 @@ name: help
 description: "Usage `rdc:help` — Show all MCP skills with usage, requirements, slash forms, and codeflow status. Reads `.claude-plugin/plugin.json` skills_meta (single source of truth). Call when unsure which skill to use or what args it takes."
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 > **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`
 > Checklist-only output. No tool-call narration. No raw MCP/JSON/log dumps.
 

package/skills/lifeai-brochure-author/SKILL.md

@@ -20,6 +20,7 @@ required_validators:
     blocking: true
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 # LIFEAI Brochure Authoring Contract
 > **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`
 > Return the authored JSX guidance or verification result directly; do not dump raw tool logs.

package/skills/new-model/SKILL.md

@@ -18,6 +18,7 @@ description: >
   Trigger when asked to create, register, or check whether a reference model exists.
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 > **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`
 > Checklist-only output. No tool-call narration. No raw MCP/JSON/log dumps.
 > One checklist upfront, updated in place, shown again at end with a 1-line verdict.

package/skills/package-design/SKILL.md

@@ -6,6 +6,7 @@ description: >-
   and sit at the right size.
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 > **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`
 > Checklist-only output. No tool-call narration. No raw MCP/JSON/log dumps.
 > One checklist upfront, updated in place, shown again at end with a 1-line verdict.

package/skills/pattern-advisor/SKILL.md

@@ -8,6 +8,7 @@ description: >-
   only; never rewrites code itself.
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 > **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`
 > Checklist-only output. No tool-call narration. No raw MCP/JSON/log dumps.
 > One checklist upfront, updated in place, shown again at end with a 1-line verdict.

package/skills/pattern-refactoring-guide/SKILL.md

@@ -6,6 +6,7 @@ description: >-
   concrete before/after refactor plan. Produces a plan, does not apply it.
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 > **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`
 > Checklist-only output. No tool-call narration. No raw MCP/JSON/log dumps.
 > One checklist upfront, updated in place, shown again at end with a 1-line verdict.

package/skills/refactor/SKILL.md

@@ -3,6 +3,7 @@ name: refactor
 description: "Usage `rdc:refactor <epic-id|topic> --takeover <reason>` — Governed consolidation and implementation of a cross-cutting refactor. Authorizes an explicit takeover, preserves evidence, re-parents relevant work through the consolidation RPC, isolates every writer, and uses the same review, validator, landing, and delivery gates as rdc:build."
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 > **OUTPUT CONTRACT:** `guides/output-contract.md`
 
 # rdc:refactor

package/skills/regen-media/SKILL.md

@@ -3,6 +3,7 @@ name: regen-media
 description: "Usage `rdc:regen-media <generate|edit|upscale|upload> <brief-or-path>` - Primary image-generation and Regen Media asset workflow. Use for image generation, image editing/upscaling, GPT Image/gpt-image-2 requests, Codex built-in image_gen, and uploading finished images to regen-media/R2. The default path is keyless local Codex gpt-image-2 via the built-in image_gen tool; server-side regen-media MCP/API generation is fallback only."
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 > **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`
 > Checklist-only output. No tool-call narration. No raw MCP/JSON/log dumps.
 > One checklist upfront, updated in place, shown again at end with a 1-line verdict.

package/skills/solid-validator/SKILL.md

@@ -8,6 +8,7 @@ description: >-
   `scripts/solid-score.mjs` for the mechanism.
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 > **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`
 > Checklist-only output. No tool-call narration. No raw MCP/JSON/log dumps.
 > One checklist upfront, updated in place, shown again at end with a 1-line verdict.

package/skills/terminal-config/SKILL.md

@@ -3,6 +3,7 @@ name: terminal-config
 description: "Usage `rdc:terminal-config <task>` — read and safely modify Windows Terminal settings, shell profiles, and agent startup sequencing without relying on machine-specific paths."
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 > **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`
 > Checklist-only output. No tool-call narration. No raw MCP/JSON/log dumps.
 > One checklist upfront, updated in place, shown again at end with a 1-line verdict.

package/skills/testing-strategy/SKILL.md

@@ -8,6 +8,7 @@ description: >-
   covers whether a surface's behavior is actually provable, and how.
 ---
 
+> If dispatching subagents or running as a subagent: read `{PROJECT_ROOT}/.rdc/guides/agent-bootstrap.md` first (fallback: `.rdc/guides/agent-bootstrap.md`) — this is also where the global rdc-harness-use policy for create/open/build/deploy work lives.
 > **⚠️ OUTPUT CONTRACT (READ FIRST):** `guides/output-contract.md`
 > Checklist-only output. No tool-call narration. No raw MCP/JSON/log dumps.
 > One checklist upfront, updated in place, shown again at end with a 1-line verdict.

package/tests/foreground-process-gate.test.mjs

@@ -0,0 +1,114 @@
+#!/usr/bin/env node
+/**
+ * foreground-process-gate.js — narrowed to Playwright headed/UI-mode only
+ * (epic 688ad6da WP-4, lifeai-env). Four window-focus/process-launch checks
+ * (checkWindowFocusApi, checkPowerShell, checkCmdStart, checkDirectShellLaunch)
+ * were removed; the replacement is lifeai-env's lib/TermLaunch.psm1 caller-
+ * logging, which lives in a different repo and is not this file's job to
+ * re-implement or re-check.
+ *
+ * This file's job is now to prove two things stay true going forward:
+ *   1. The four retired checks no longer block anything -- a real regression
+ *      here would silently reintroduce ceremony an operator explicitly
+ *      retired, with no other test anywhere positioned to catch it.
+ *   2. checkPlaywright survives untouched (Design Decision D3) -- this is a
+ *      SEPARATE safety property, not part of the terminal-launch
+ *      consolidation, and narrowing the file must never have narrowed this.
+ *
+ * Run: node tests/foreground-process-gate.test.mjs
+ */
+import { dirname, join, resolve } from 'node:path';
+import { fileURLToPath } from 'node:url';
+import { spawnSync } from 'node:child_process';
+
+const __dirname = dirname(fileURLToPath(import.meta.url));
+const REPO_ROOT = resolve(__dirname, '..');
+const HOOK = join(REPO_ROOT, 'hooks', 'foreground-process-gate.js');
+
+const failures = [];
+function assert(name, condition, detail = '') {
+  if (!condition) failures.push(`${name}${detail ? `: ${detail}` : ''}`);
+  else process.stdout.write(`  ok  ${name}\n`);
+}
+
+function runGate(command) {
+  const res = spawnSync(process.execPath, [HOOK], {
+    input: JSON.stringify({ tool_name: 'Bash', tool_input: { command } }),
+    encoding: 'utf8',
+  });
+  const blocked = res.status === 1;
+  return { blocked, stdout: res.stdout, status: res.status };
+}
+
+// ===========================================================================
+// 1. The four retired checks no longer block anything.
+// ===========================================================================
+
+{
+  const r = runGate('Start-Process -FilePath node.exe -ArgumentList "server.js"');
+  assert('Start-Process with no -WindowStyle Hidden no longer blocks (checkPowerShell retired)',
+    r.blocked === false, `status=${r.status} stdout=${r.stdout}`);
+}
+
+{
+  const r = runGate('cmd /c start node.exe server.js');
+  assert('cmd /c start without /min no longer blocks (checkCmdStart retired)',
+    r.blocked === false, `status=${r.status} stdout=${r.stdout}`);
+}
+
+{
+  const r = runGate('powershell.exe -File C:\\Dev\\some-script.ps1');
+  assert('bare PowerShell .ps1 launch no longer blocks (checkDirectShellLaunch retired)',
+    r.blocked === false, `status=${r.status} stdout=${r.stdout}`);
+}
+
+{
+  const r = runGate('[Win32]::SetForegroundWindow($handle)');
+  assert('a window-focus Win32 API call no longer blocks (checkWindowFocusApi retired)',
+    r.blocked === false, `status=${r.status} stdout=${r.stdout}`);
+}
+
+// ===========================================================================
+// 2. checkPlaywright survives, untouched -- a separate safety property.
+// ===========================================================================
+
+{
+  const r = runGate('npx playwright test --headed');
+  assert('Playwright --headed still blocks',
+    r.blocked === true, `status=${r.status} stdout=${r.stdout}`);
+  assert('Playwright --headed block message names the right reason',
+    /must run headless/.test(r.stdout), r.stdout);
+}
+
+{
+  const r = runGate('PWDEBUG=1 npx playwright test');
+  assert('Playwright PWDEBUG=1 still blocks',
+    r.blocked === true, `status=${r.status} stdout=${r.stdout}`);
+}
+
+{
+  const r = runGate('npx playwright show-report');
+  assert('Playwright show-report still blocks',
+    r.blocked === true, `status=${r.status} stdout=${r.stdout}`);
+  assert('Playwright show-report block message names the right reason',
+    /launches foreground UI/.test(r.stdout), r.stdout);
+}
+
+{
+  const r = runGate('npx playwright test --reporter=list');
+  assert('an ordinary headless Playwright run is NOT blocked',
+    r.blocked === false, `status=${r.status} stdout=${r.stdout}`);
+}
+
+{
+  const r = runGate('git status');
+  assert('an unrelated ordinary command passes clean', r.blocked === false, `status=${r.status} stdout=${r.stdout}`);
+}
+
+// ===========================================================================
+if (failures.length > 0) {
+  console.error('\nforeground-process-gate tests — FAIL\n');
+  for (const f of failures) console.error(`  - ${f}`);
+  process.exit(1);
+}
+console.log('\nforeground-process-gate tests — PASS');

package/tests/harness-gates.test.mjs

@@ -146,34 +146,16 @@ const WI = 'aaaaaaaa-bbbb-cccc-dddd-eeeeeeeeeeee';
 }
 
 // ===========================================================================
-// 2. foreground-process-gate.js
+// 2. foreground-process-gate.js — moved to its own file, tests/foreground-process-gate.test.mjs
+//
+// Narrowed 2026-08-26 (epic 688ad6da WP-4, lifeai-env): the window-focus/
+// process-launch checks this section used to assert (SetForegroundWindow
+// blocking, hidden/minimized allowances) were retired — that whole class is
+// now caller-logged by lifeai-env's lib/TermLaunch.psm1, not hard-blocked
+// here. The dedicated test file covers what remains (checkPlaywright) plus
+// asserts the four retired checks stay retired, so removing this embedded
+// section is not a coverage loss — see that file instead.
 // ===========================================================================
-{
-  const focusPayload = {
-    tool_input: {
-      command: "powershell -NoProfile -Command \"Add-Type '[DllImport(\\\"user32.dll\\\")] public static extern bool SetForegroundWindow(System.IntPtr hWnd);'\"",
-    },
-  };
-  const r = runHook('foreground-process-gate.js', focusPayload, {});
-  assert('FPG blocks SetForegroundWindow focus API', r.status === 1, `status=${r.status} ${r.stdout}${r.stderr}`);
-  assert('FPG block mentions window focus operations', /Window focus\/restore\/minimize\/collapse/.test(r.stdout + r.stderr));
-
-  const hiddenPayload = {
-    tool_input: {
-      command: 'powershell.exe -NoProfile -NonInteractive -WindowStyle Hidden -File ".\\\\scripts\\\\helper.ps1"',
-    },
-  };
-  const h = runHook('foreground-process-gate.js', hiddenPayload, {});
-  assert('FPG allows hidden PowerShell helper', h.status === 0, `status=${h.status} ${h.stdout}${h.stderr}`);
-
-  const minimizedPayload = {
-    tool_input: {
-      command: 'Start-Process powershell.exe -WindowStyle Minimized -ArgumentList "-NoProfile"',
-    },
-  };
-  const m = runHook('foreground-process-gate.js', minimizedPayload, {});
-  assert('FPG allows minimized Start-Process without focus APIs', m.status === 0, `status=${m.status} ${m.stdout}${m.stderr}`);
-}
 
 // ===========================================================================
 // 3. post-tool-batch-gate.js