fix(roster): point all companion-skills at broomva/skills + drift-guard (BRO-1584)

.github/workflows/ci.yml

@@ -83,6 +83,12 @@ jobs:
         run: python3 -m pip install --quiet PyYAML
 
       - name: Run all tests/*.test.sh
+        env:
+          # Lets roster-monorepo-sync.test.sh run its remote resolution check
+          # (every roster name resolves in broomva/skills). Skips gracefully if the
+          # tree can't be fetched (rate-limit / private); the offline structural
+          # checks always run regardless.
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         # As of v0.8.0 (Phase 5 of substrate completion) the pre-existing
         # tests/template_lockstep.test.sh and tests/onboard.test.sh are
         # fixed (case-insensitive lockstep assertion + BSTACK_SKIP_SKILLS=1

bin/bstack-skills

@@ -249,8 +249,16 @@ cmd_install() {
             fi
         fi
         echo "  [install] $n ($r)..."
+        # broomva/skills is the monorepo (many skills) — install the single skill
+        # with --skill <name>, else a bare `add broomva/skills` pulls everything.
+        # Standalone repos install whole.
+        if [ "$r" = "broomva/skills" ]; then
+            add_args=("$r" --skill "$n")
+        else
+            add_args=("$r")
+        fi
         # shellcheck disable=SC2086
-        if $NPX_CMD "$r" >/dev/null 2>&1; then
+        if $NPX_CMD "${add_args[@]}" >/dev/null 2>&1; then
             installed=$((installed + 1))
             echo "    ✓ installed"
         else

references/companion-skills.yaml

@@ -17,22 +17,22 @@ schema_version: 1
 skills:
   # ── Foundation: orchestration + governance modes ────────────────────────
   - name: autonomous
-    repo: broomva/autonomous
+    repo: broomva/skills
     category: orchestration
     required: true
     introduced_in: 0.1.0
     description: "Canonical operating mode that fires the 20-reflex pipeline"
 
   - name: cross-review
-    repo: broomva/cross-review
+    repo: broomva/skills
     category: safety
     primitive: P20
     required: true
     introduced_in: 0.2.0
     description: "Cross-model adversarial review gate"
 
   - name: role-x
-    repo: broomva/role-x
+    repo: broomva/skills
     category: meta
     primitive: P17
     required: true
@@ -41,19 +41,12 @@ skills:
 
   # ── Control + knowledge stack ───────────────────────────────────────────
   - name: agentic-control-kernel
-    repo: broomva/agentic-control-kernel
+    repo: broomva/skills
     category: meta
     required: true
     introduced_in: 0.1.0
     description: "Unifying control-systems metalayer kernel"
 
-  - name: control-metalayer-loop
-    repo: broomva/control-metalayer-loop
-    category: meta
-    required: true
-    introduced_in: 0.1.0
-    description: "Governance feedback loop"
-
   - name: harness-engineering-playbook
     repo: broomva/skills
     skillPath: skills/harness-engineering-playbook/SKILL.md
@@ -63,23 +56,8 @@ skills:
     min_bstack_version: 0.21.0
     description: "Patterns for agent harness design — AGENTS.md, smoke/test/lint/typecheck harness, entropy-control checks (migrated 2026-05-25 from broomva/harness-engineering-skill — Phase 4c; renamed)"
 
-  - name: agent-consciousness
-    repo: broomva/agent-consciousness
-    category: knowledge
-    required: true
-    introduced_in: 0.1.0
-    description: "Three-substrate persistence + episodic memory"
-
-  - name: knowledge-graph-memory
-    repo: broomva/knowledge-graph-memory
-    category: knowledge
-    primitive: P6
-    required: true
-    introduced_in: 0.1.0
-    description: "Conversation logs → Obsidian-shaped knowledge graph"
-
   - name: kg
-    repo: broomva/kg
+    repo: broomva/skills
     category: knowledge
     primitive: P6
     required: true
@@ -88,42 +66,21 @@ skills:
     description: "LLM-as-index loader for research/entities/. Two-tier scoring (catalog tier-1, body-grep tier-2 fallback) routes a topic to top-N entity bodies the agent reasons over. Pairs with bookkeeping's cmd_index (catalog generator) and the workspace's knowledge-catalog-refresh-hook.sh Stop hook (P6)."
 
   - name: prompt-library
-    repo: broomva/prompt-library
+    repo: broomva/skills
     category: meta
     required: true
     introduced_in: 0.1.0
     description: "Prompt management CLI + API"
 
   # ── Productive-wait + orchestration ─────────────────────────────────────
   - name: p9
-    repo: broomva/p9
+    repo: broomva/skills
     category: lifecycle
     primitive: P9
     required: true
     introduced_in: 0.1.0
     description: "CI watcher + productive-wait"
 
-  - name: symphony
-    repo: broomva/symphony
-    category: orchestration
-    required: true
-    introduced_in: 0.1.0
-    description: "Orchestration daemon"
-
-  - name: symphony-forge
-    repo: broomva/symphony-forge
-    category: orchestration
-    required: true
-    introduced_in: 0.1.0
-    description: "Project scaffolding with control metalayer"
-
-  - name: autoany
-    repo: broomva/autoany
-    category: meta
-    required: true
-    introduced_in: 0.1.0
-    description: "EGRI L2 meta-controller (self-improvement)"
-
   - name: alkosto-wait-optimizer
     repo: broomva/skills
     skillPath: skills/alkosto-wait-optimizer/SKILL.md
@@ -191,15 +148,6 @@ skills:
     min_bstack_version: 0.21.0
     description: "Deep-ocean genomics research — eDNA, AlphaFold, ESMFold, Evo 2, marine biodiversity, agentic bioinformatics"
 
-  - name: microgrid-agent
-    repo: broomva/skills
-    skillPath: skills/microgrid-agent/SKILL.md
-    category: platform
-    required: false
-    introduced_in: 0.21.4
-    min_bstack_version: 0.21.0
-    description: "Edge AI agent for autonomous renewable energy microgrid management on Raspberry Pi"
-
   # Compute & remote infrastructure
   - name: colab-remote
     repo: broomva/skills
@@ -343,35 +291,21 @@ skills:
     description: "In-silico neuroscience experiments using Meta's TRIBE v2 — predict fMRI cortical responses to video, audio & text without brain scanners"
 
   # ── Discovery + skill management ────────────────────────────────────────
-  - name: skills
+  - name: skills-catalog
     repo: broomva/skills
     category: meta
     required: true
     introduced_in: 0.1.0
     description: "Canonical reference of all available skills"
 
-  - name: skills-showcase
-    repo: broomva/skills-showcase
-    category: meta
-    required: false
-    introduced_in: 0.1.0
-    description: "Skill catalog presentation + Remotion showcase"
-
   # ── Design + frontend ───────────────────────────────────────────────────
   - name: arcan-glass
-    repo: broomva/arcan-glass
+    repo: broomva/skills
     category: design
     required: false
     introduced_in: 0.1.0
     description: "Broomva trademark web styling system"
 
-  - name: next-forge
-    repo: broomva/next-forge
-    category: platform
-    required: false
-    introduced_in: 0.1.0
-    description: "Next.js project scaffolding"
-
   - name: brand-icons
     repo: broomva/skills
     skillPath: skills/brand-icons/SKILL.md
@@ -566,7 +500,7 @@ skills:
 
   # ── Platform + finance ──────────────────────────────────────────────────
   - name: finance-substrate
-    repo: broomva/finance-substrate
+    repo: broomva/skills
     category: platform
     required: false
     introduced_in: 0.1.0

references/skills-roster.md

@@ -2,6 +2,8 @@
 
 28 curated skills across 7 layers. The Broomva Stack.
 
+> **Canonical *installable* roster:** [`references/companion-skills.yaml`](companion-skills.yaml) — the 56 bstack-native skills, all in the **broomva/skills** monorepo (validated by `tests/roster-monorepo-sync.test.sh`). Install path-independently: `npx skills add broomva/skills --skill <name>`. **This doc is a broader human catalog**: it also documents ecosystem *products* (e.g. symphony, autoany, next-forge) that are not installable bstack skills, so it intentionally lists more than the YAML. For what is installable, the YAML is authoritative.
+
 ## Publishing standard (skill layout) — required for installability
 
 A skill is a **folder**, per the [Agent Skills standard](https://agentskills.io) — `SKILL.md` + optional `scripts/` / `references/` / `assets/`. **Publish every multi-file skill in a `skills/<name>/` subdirectory of its repo — never as a bare top-level `SKILL.md` at the repo root.**
@@ -19,26 +21,26 @@ A skill is a **folder**, per the [Agent Skills standard](https://agentskills.io)
 
 | # | Skill | Install | Description |
 |---|-------|---------|-------------|
-| 1 | `agentic-control-kernel` | `npx skills add broomva/agentic-control-kernel` | LLM-as-controller with safety shields, typed plant/action/trace schemas, multi-rate loop hierarchy. The governance backbone. |
-| 2 | `control-metalayer-loop` | `npx skills add broomva/control-metalayer` | Control primitives: setpoints, sensors, actuators, stability gates, policy profiles. Bootstraps `.control/policy.yaml`. |
+| 1 | `agentic-control-kernel` | `npx skills add broomva/skills --skill agentic-control-kernel` | LLM-as-controller with safety shields, typed plant/action/trace schemas, multi-rate loop hierarchy. The governance backbone. |
+| 2 | `control-metalayer-loop` | *(consolidated → `agentic-control-kernel`)* | Control primitives: setpoints, sensors, actuators, stability gates, policy profiles. Bootstraps `.control/policy.yaml`. |
 | 3 | `harness-engineering-playbook` | `npx skills add broomva/skills --skill harness-engineering-playbook` | Agent-first workflow: AGENTS.md, smoke/test/lint/typecheck harness, entropy-control checks. (Migrated 2026-05-25 from `broomva/harness-engineering-skill` — Phase 4c.) |
 
 ## Memory & Consciousness
 
 | # | Skill | Install | Description |
 |---|-------|---------|-------------|
-| 4 | `agent-consciousness` | `npx skills add broomva/control-metalayer` | Three-substrate persistence: governance + knowledge graph + episodic memory. Progressive crystallization pathway. |
-| 5 | `knowledge-graph-memory` | `npx skills add broomva/control-metalayer` | Conversation logs to Obsidian knowledge graph bridge. Generates per-session docs with frontmatter and wikilinks. |
-| 6 | `kg` | `npx skills add broomva/kg` | LLM-as-index loader for `research/entities/`. Two-tier scoring (catalog tier-1, body-grep tier-2 fallback) routes a topic to top-N entity bodies the agent reasons over. The runtime form of "the LLM **is** the index" — substrate canonical, one projection (catalog) routes, agent IS the query engine. Pairs with bookkeeping's `cmd_index` and the workspace's `knowledge-catalog-refresh-hook.sh` Stop hook (P6). |
-| 7 | `prompt-library` | `npx skills add broomva/prompt-library` | Shared knowledge surface across agents. Versioned prompts + an evaluation engine: every pull/completion writes a typed `prompt_invocation` row with source attribution (`web\|cli\|skill\|api`), latency, tokens, cost, and feedback. The `broomva` Rust CLI (`broomva prompts pull/list/complete/feedback`) is the runtime; the broomva-cli skill carries the auto-tracing mandate. See [prompts-integration.md](prompts-integration.md). |
+| 4 | `agent-consciousness` | *(consolidated → `agentic-control-kernel`)* | Three-substrate persistence: governance + knowledge graph + episodic memory. Progressive crystallization pathway. |
+| 5 | `knowledge-graph-memory` | *(consolidated → `agentic-control-kernel`)* | Conversation logs to Obsidian knowledge graph bridge. Generates per-session docs with frontmatter and wikilinks. |
+| 6 | `kg` | `npx skills add broomva/skills --skill kg` | LLM-as-index loader for `research/entities/`. Two-tier scoring (catalog tier-1, body-grep tier-2 fallback) routes a topic to top-N entity bodies the agent reasons over. The runtime form of "the LLM **is** the index" — substrate canonical, one projection (catalog) routes, agent IS the query engine. Pairs with bookkeeping's `cmd_index` and the workspace's `knowledge-catalog-refresh-hook.sh` Stop hook (P6). |
+| 7 | `prompt-library` | `npx skills add broomva/skills --skill prompt-library` | Shared knowledge surface across agents. Versioned prompts + an evaluation engine: every pull/completion writes a typed `prompt_invocation` row with source attribution (`web\|cli\|skill\|api`), latency, tokens, cost, and feedback. The `broomva` Rust CLI (`broomva prompts pull/list/complete/feedback`) is the runtime; the broomva-cli skill carries the auto-tracing mandate. See [prompts-integration.md](prompts-integration.md). |
 
 ## Orchestration
 
 | # | Skill | Install | Description |
 |---|-------|---------|-------------|
-| 8 | `symphony` | `npx skills add broomva/symphony` | Rust orchestration engine for coding agents. Daemon mode, Linear/GitHub tracker integration, lifecycle hooks. |
-| 9 | `symphony-forge` | `npx skills add broomva/symphony-forge` | CLI scaffolder with composable control metalayer. Bootstraps projects with agent governance built in. |
-| 10 | `autoany` | `npx skills add broomva/autoany` | EGRI self-improvement framework. Turns ambiguous goals into safe, measurable, rollback-capable recursive improvement loops. |
+| 8 | `symphony` | *(ecosystem product — not a portable skill in the monorepo; lives at `broomva/symphony`)* | Rust orchestration engine for coding agents. Daemon mode, Linear/GitHub tracker integration, lifecycle hooks. |
+| 9 | `symphony-forge` | *(ecosystem product — not a portable skill in the monorepo; lives at `broomva/symphony-forge`)* | CLI scaffolder with composable control metalayer. Bootstraps projects with agent governance built in. |
+| 10 | `autoany` | *(ecosystem product — not a portable skill in the monorepo; lives at `broomva/autoany`)* | EGRI self-improvement framework. Turns ambiguous goals into safe, measurable, rollback-capable recursive improvement loops. |
 
 ## Research & Intelligence
 
@@ -47,27 +49,27 @@ A skill is a **folder**, per the [Agent Skills standard](https://agentskills.io)
 | 11 | `deep-dive-research-orchestrator` | `npx skills add broomva/skills --skill deep-dive-research-orchestrator` | Multi-dimensional research with coordinated AI specialists. 10+ source synthesis with citations. (Migrated 2026-05-25 from `broomva/deep-dive-research-skill` — Phase 4c.) |
 | 11a | `social-intelligence` | `npx skills add broomva/skills --skill social-intelligence` | Autonomous social engagement + knowledge extraction loop for Moltbook and X — compounds with `blog-post` and `content-creation`. (Migrated 2026-05-25 from `broomva/social-intelligence` — Phase 4c.) |
 | 12 | `skills` | `npx skills add broomva/skills` | Canonical reference inventory of 83 agent skills across 15 domains. Browsable catalog. |
-| 13 | `skills-showcase` | `npx skills add broomva/skills` | Remotion video + X thread generator for the skills inventory. Animated showcase content. |
+| 13 | `skills-showcase` | *(repo-root Remotion tool in broomva/skills — not an installable skill)* | Remotion video + X thread generator for the skills inventory. Animated showcase content. |
 
 ## Design & Implementation
 
 | # | Skill | Install | Description |
 |---|-------|---------|-------------|
-| 14 | `arcan-glass` | `npx skills add broomva/arcan-glass` | BroomVA web design system. Glass/frosted effects, dark-first themes, AI Blue brand tokens. |
-| 15 | `next-forge` | `npx skills add broomva/symphony-forge` | Production Next.js SaaS template via symphony-forge. Turborepo, auth, payments, observability. |
+| 14 | `arcan-glass` | `npx skills add broomva/skills --skill arcan-glass` | BroomVA web design system. Glass/frosted effects, dark-first themes, AI Blue brand tokens. |
+| 15 | `next-forge` | *(ecosystem product — not a portable skill in the monorepo; lives at `broomva/symphony-forge`)* | Production Next.js SaaS template via symphony-forge. Turborepo, auth, payments, observability. |
 
 ## Platform
 
 | # | Skill | Install | Description |
 |---|-------|---------|-------------|
-| 16 | `alkosto-wait-optimizer` | `npx skills add broomva/alkosto-wait-optimizer-skill` | Probability-based decision tool for optimal waiting times. Bayesian estimation with uncertainty. |
+| 16 | `alkosto-wait-optimizer` | `npx skills add broomva/skills --skill alkosto-wait-optimizer` | Probability-based decision tool for optimal waiting times. Bayesian estimation with uncertainty. |
 | 17 | `content-creation` | `npx skills add broomva/skills --skill content-creation` | Full-stack content pipeline — research → narrative → visual assets → video → social → deploy; ships bstack-launch + open-source-stack example campaigns. (Migrated 2026-05-25 from `broomva/content-creation` — Phase 4b.) |
 | 17a | `content-engine` | `npx skills add broomva/skills --skill content-engine` | Full-stack AI content studio — visual DNA compiler, cinematic generation, browser autopilot, content loop; bundles 4 sub-skills. (Migrated 2026-05-25 from `broomva/content-engine` — Phase 4b.) |
 | 17b | `launch-video` | `npx skills add broomva/skills --skill launch-video` | Liquid Glass product launch video — dark void, 3D floating panels, spring animations via Remotion. (Migrated 2026-05-25 from `broomva/launch-video` — Phase 4b.) |
 | 17c | `ltx-video` | `npx skills add broomva/skills --skill ltx-video` | LTX-2.3 video generation — setup, inference, prompting, ComfyUI integration for Lightricks 22B DiT audio-video model. (Migrated 2026-05-25 from `broomva/ltx-video` — Phase 4b.) |
 | 17d | `creative-review` | `npx skills add broomva/skills --skill creative-review` | Meta creative review — style adherence scoring, feedback loops, self-improving creative pipeline. (Migrated 2026-05-25 from `broomva/creative-review` — Phase 4b.) |
 | 17e | `brainrot-for-good` | `npx skills add broomva/skills --skill brainrot-for-good` | High-retention video production using dopamine-aware editing for genuinely valuable content. (Migrated 2026-05-25 from `broomva/brainrot-for-good` — Phase 4b.) |
-| 18 | `finance-substrate` | `npx skills add broomva/finance-substrate` | Personal finance & Colombian tax management. Bank CSV import, TRM rates, DIAN tax projection, withholdings, e-invoicing. Zero paid deps. *(Pending Tier-1 vs Tier-2 lock-in.)* |
+| 18 | `finance-substrate` | `npx skills add broomva/skills --skill finance-substrate` | Personal finance & Colombian tax management. Bank CSV import, TRM rates, DIAN tax projection, withholdings, e-invoicing. Zero paid deps. *(Pending Tier-1 vs Tier-2 lock-in.)* |
 | 18a | `investment-management` | `npx skills add broomva/skills --skill investment-management` | Portfolio construction, factor models, backtesting, multi-platform execution (Alpaca, Coinbase, Polymarket). (Migrated 2026-05-25 from `broomva/investment-management` — Phase 4c.) |
 | 18b | `wealth-management` | `npx skills add broomva/skills --skill wealth-management` | Wealth planning + Monte Carlo simulations + tax-optimized allocation + net worth forecasting. (Migrated 2026-05-25 from `broomva/wealth-management` — Phase 4c.) |
 | 18c | `haima` | `npx skills add broomva/skills --skill haima` | Agent guide for x402 machine-to-machine payments, secp256k1 wallets, per-task billing, on-chain USDC settlement. (Migrated 2026-05-25 from `broomva/haima-skill` — Phase 4c; renamed to drop `-skill` suffix. Runtime crate stays at `broomva/haima`.) |