garrytan · ljagged · Mar 26, 2026 · Mar 26, 2026 · Mar 26, 2026 · Mar 26, 2026
diff --git a/.gitignore b/.gitignore
@@ -17,3 +17,4 @@ extension/.auth.json
 .env.*
 !.env.example
 supabase/.temp/
+SRD
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,5 +1,23 @@
 # Changelog
 
+## [0.13.4.0] - 2026-03-28 — Build Your Social Strategy
+
+New `/social-strategy` skill helps founders build authentic public presence. Not a content mill. Not a social media manager. The strategic scaffolding that makes your own content effective.
+
+### Added
+
+- **`/social-strategy audit`** researches your competitive social landscape. Maps competitor social activity, identifies 10-15 thought leaders with actionable engagement URLs, recommends which platforms to focus on (and which to ignore), and generates voice prompt writing exercises calibrated to your audit findings.
+- **`/social-strategy` session** walks you through a 6-step interactive strategy. Defines your voice fingerprint (qualitative, not scored), builds content pillars grounded in your competitive gaps, maps relationship priorities, locks in a realistic platform plan and cadence, and aligns content themes to product milestones.
+- **`/social-strategy refresh`** is a lightweight competitive re-scan. Checks what changed since your last audit without re-running the full research. Writes a new timestamped doc (never mutates the original).
+- **Voice prompts with intentional temporal gap.** The audit generates writing exercises the founder fills between sessions, not inline. Better signal, better fingerprints.
+- **Social strategy context in CEO reviews.** `/plan-ceo-review` now discovers social strategy docs and surfaces content pillars and platform recommendations during scope decisions.
+- **3 Architecture Decision Records** (docs/adrs/0003-0005) documenting the voice fingerprinting approach, framework panel design, and the faithful representation requirement.
+
+### For contributors
+
+- New routing test for `/social-strategy` in `test/skill-routing-e2e.test.ts`
+- Touchfile entry for diff-based test selection
+
 ## [0.13.3.0] - 2026-03-28 — Lock It Down
 
 Six fixes from community PRs and bug reports. The big one: your dependency tree is now pinned. Every `bun install` resolves the exact same versions, every time. No more floating ranges pulling fresh packages from npm on every setup.
@@ -105,6 +123,7 @@ Fixes 20 Socket alerts and 3 Snyk findings from the skills.sh security audit. Yo
 ### For contributors
 
 - New `test:audit` script runs 6 regression tests that enforce all audit fixes stay in place.
+>>>>>>> upstream/main
 
 ## [0.12.11.0] - 2026-03-27 — Skill Prefix is Now Your Choice
 

diff --git a/README.md b/README.md
@@ -46,7 +46,7 @@ Fork it. Improve it. Make it yours. And if you want to hate on free open source
 
 Open Claude Code and paste this. Claude does the rest.
 
-> Install gstack: run **`git clone --single-branch --depth 1 https://github.com/garrytan/gstack.git ~/.claude/skills/gstack && cd ~/.claude/skills/gstack && ./setup`** then add a "gstack" section to CLAUDE.md that says to use the /browse skill from gstack for all web browsing, never use mcp\_\_claude-in-chrome\_\_\* tools, and lists the available skills: /office-hours, /plan-ceo-review, /plan-eng-review, /plan-design-review, /design-consultation, /design-shotgun, /review, /ship, /land-and-deploy, /canary, /benchmark, /browse, /connect-chrome, /qa, /qa-only, /design-review, /setup-browser-cookies, /setup-deploy, /retro, /investigate, /document-release, /codex, /cso, /autoplan, /careful, /freeze, /guard, /unfreeze, /gstack-upgrade. Then ask the user if they also want to add gstack to the current project so teammates get it.
+> Install gstack: run **`git clone --single-branch --depth 1 https://github.com/garrytan/gstack.git ~/.claude/skills/gstack && cd ~/.claude/skills/gstack && ./setup`** then add a "gstack" section to CLAUDE.md that says to use the /browse skill from gstack for all web browsing, never use mcp\_\_claude-in-chrome\_\_\* tools, and lists the available skills: /office-hours, /plan-ceo-review, /plan-eng-review, /plan-design-review, /design-consultation, /design-shotgun, /review, /ship, /land-and-deploy, /canary, /benchmark, /browse, /connect-chrome, /qa, /qa-only, /design-review, /setup-browser-cookies, /setup-deploy, /retro, /investigate, /document-release, /codex, /cso, /autoplan, /strategist, /social-strategy, /careful, /freeze, /guard, /unfreeze, /gstack-upgrade. Then ask the user if they also want to add gstack to the current project so teammates get it.
 
 ### Step 2: Add to your repo so teammates get it (optional)
 
@@ -162,6 +162,8 @@ Each skill feeds into the next. `/office-hours` writes a design doc that `/plan-
 | `/canary` | **SRE** | Post-deploy monitoring loop. Watches for console errors, performance regressions, and page failures. |
 | `/benchmark` | **Performance Engineer** | Baseline page load times, Core Web Vitals, and resource sizes. Compare before/after on every PR. |
 | `/document-release` | **Technical Writer** | Update all project docs to match what you just shipped. Catches stale READMEs automatically. |
+| `/strategist` | **Senior Strategist** | Competitive strategy with framework orchestration. Brief mode: autonomous competitive intelligence with cited research. Session mode: Rumelt's kernel (diagnosis → guiding policy → coherent actions → milestone-gated execution). |
+| `/social-strategy` | **Communications Strategist** | Build your public presence. Audit mode: research competitor social activity, map thought leaders with actionable URLs, recommend platforms. Session mode: define your voice, content pillars, relationship priorities, and realistic cadence. Never generates content to post. |
 | `/retro` | **Eng Manager** | Team-aware weekly retro. Per-person breakdowns, shipping streaks, test health trends, growth opportunities. `/retro global` runs across all your projects and AI tools (Claude Code, Codex, Gemini). |
 | `/browse` | **QA Engineer** | Give the agent eyes. Real Chromium browser, real clicks, real screenshots. ~100ms per command. `$B connect` launches your real Chrome as a headed window — watch every action live. |
 | `/setup-browser-cookies` | **Session Manager** | Import cookies from your real browser (Chrome, Arc, Brave, Edge) into the headless session. Test authenticated pages. |
@@ -277,7 +279,7 @@ Use /browse from gstack for all web browsing. Never use mcp__claude-in-chrome__*
 Available skills: /office-hours, /plan-ceo-review, /plan-eng-review, /plan-design-review,
 /design-consultation, /review, /ship, /land-and-deploy, /canary, /benchmark, /browse,
 /qa, /qa-only, /design-review, /setup-browser-cookies, /setup-deploy, /retro,
-/investigate, /document-release, /codex, /cso, /autoplan, /careful, /freeze, /guard,
+/investigate, /document-release, /codex, /cso, /autoplan, /strategist, /careful, /freeze, /guard,
 /unfreeze, /gstack-upgrade.
 ```
 

diff --git a/SKILL.md b/SKILL.md
@@ -265,7 +265,8 @@ Only run skills the user explicitly invokes. This preference persists across ses
 If `PROACTIVE` is `true` (default): suggest adjacent gstack skills when relevant to the
 user's workflow stage:
 - Brainstorming → /office-hours
-- Strategy → /plan-ceo-review
+- Strategy → /plan-ceo-review or /strategist
+- Competitive analysis → /strategist
 - Architecture → /plan-eng-review
 - Design → /plan-design-review or /design-consultation
 - Auto-review → /autoplan

diff --git a/SKILL.md.tmpl b/SKILL.md.tmpl
@@ -23,7 +23,8 @@ Only run skills the user explicitly invokes. This preference persists across ses
 If `PROACTIVE` is `true` (default): suggest adjacent gstack skills when relevant to the
 user's workflow stage:
 - Brainstorming → /office-hours
-- Strategy → /plan-ceo-review
+- Strategy → /plan-ceo-review or /strategist
+- Competitive analysis → /strategist
 - Architecture → /plan-eng-review
 - Design → /plan-design-review or /design-consultation
 - Auto-review → /autoplan

diff --git a/TODOS.md b/TODOS.md
@@ -632,6 +632,94 @@ Shipped in v0.6.5. TemplateContext in gen-skill-docs.ts bakes skill name into pr
 **Priority:** P3
 **Depends on:** Telemetry data showing freeze hook fires in real /investigate sessions
 
+## Strategist
+
+### CSO/retro artifact ingestion
+
+**What:** Add CSO and retro artifact ingestion to `/strategist brief` when those skills gain persistent project-scoped output.
+
+**Why:** Security posture (from `/cso`) and shipping velocity (from `/retro`) are valuable strategic inputs. Currently those skills write to stdout or `.context/` dirs that aren't accessible cross-skill.
+
+**Context:** `/strategist` design doc lists these as "reads from" but they don't actually persist to `~/.gstack/projects/` yet. When `/cso` and `/retro` gain project-scoped artifact output, add the appropriate globs to `/strategist`'s context ingestion phase.
+
+**Effort:** S (once upstream skills persist artifacts)
+**Priority:** P2
+**Depends on:** `/cso` and `/retro` writing to `~/.gstack/projects/`
+
+### Autoplan integration
+
+**What:** Integrate `/strategist` into `/autoplan`'s review pipeline. If a strategy doc exists, surface it as context during reviews. Optionally offer to run `/strategist brief` first.
+
+**Why:** Without this, the default planning path (`/autoplan`) bypasses competitive strategy entirely. A strategy doc should inform scope and ambition decisions in CEO review.
+
+**Context:** `/autoplan` currently runs CEO, design, and eng reviews. Strategy docs at `*-strategy-*.md` should be discoverable alongside design docs. The simplest integration: `/autoplan` reads strategy docs if they exist (same as it reads design docs) and mentions `/strategist` as a prerequisite option.
+
+**Effort:** S
+**Priority:** P2
+**Depends on:** `/strategist` skill existing and being dogfooded
+
+### E2E eval test for /strategist
+
+**What:** Write E2E eval test after the first dogfood session. Use real output as ground truth. Test: (a) brief produces a file with inline citations, (b) session produces a strategy doc with framework selection rationale, (c) 90-day plan items are specific not generic.
+
+**Why:** Without an eval, quality regressions in the template are invisible. The template's framework selection logic and citation requirements need automated validation.
+
+**Context:** Classify as `periodic` tier (non-deterministic, quality benchmark). Dogfood session produces real output to calibrate eval expectations. Add touchfile entry in `test/helpers/touchfiles.ts` with `strategist/**` dependency.
+
+**Effort:** M
+**Priority:** P2
+**Depends on:** First dogfood session completing
+
+## Social Strategy
+
+### Framework panel curation (PREREQUISITE)
+
+**What:** Research and curate the narrative framework panel for `/social-strategy`. For each candidate thinker: engage with their actual body of work (not summaries), identify which specific ideas apply to founder content strategy, justify inclusion (what failure mode does this thinker uniquely catch), and define how the skill should represent their work faithfully.
+
+**Why:** Blocks three features: framework orientation session step, framework-annotated content pillars, and review mode audience value criterion. See ADR 0005.
+
+**Context:** Candidates from design conversation: Godin, Sierra, Miller, Cialdini, Christensen (JTBD only). Unevaluated: Handley, Schwartz, Moore (may belong in /strategist). Deliverable: a framework panel document with 4-6 thinkers, each with name, relevant works, which ideas apply, core question(s), what failure mode they catch, and justification.
+
+**Effort:** M (human: ~2 weeks reading/research, CC cannot do this — ADR 0005 explicitly requires human engagement with source material)
+**Priority:** P1
+**Depends on:** Nothing — can start immediately
+
+### Review mode (v1.1)
+
+**What:** Add `/social-strategy review` subcommand. Draft critique against voice doc + content pillars + style guide. Scores 1-10 on 5 criteria (voice consistency, pillar alignment, audience value, effectiveness habits, domain safety). Never rewrites, only critiques.
+
+**Why:** Closes the feedback loop. Without review mode, the founder has strategy artifacts but no structured way to evaluate their own content against them.
+
+**Context:** Depends on voice fingerprint quality being validated via dogfood (are fingerprints specific enough to score against?) and framework panel being curated (for audience value criterion). Pattern: inline critique only, no file output.
+
+**Effort:** M (human: ~2 weeks / CC: ~1-2 hours)
+**Priority:** P1
+**Depends on:** Framework panel curation, voice fingerprint quality validation from dogfood
+
+### E2E eval test for /social-strategy
+
+**What:** Write E2E eval test using dogfood output as ground truth. Test: (a) audit produces file with inline citations, (b) voice prompts are topically relevant to audit findings, (c) strategy doc has pillar-to-strategy linkage, (d) style guide has both effectiveness patterns and domain landmines personalized to the founder's space.
+
+**Why:** Without an eval, quality regressions in the template are invisible. Classify as periodic tier (non-deterministic, quality benchmark).
+
+**Context:** Add touchfile entry in `test/helpers/touchfiles.ts` with `social-strategy/**` dependency. Pattern: same as `/strategist` E2E eval TODO.
+
+**Effort:** M (human: ~2 weeks / CC: ~30 min)
+**Priority:** P2
+**Depends on:** First dogfood session completing
+
+### Autoplan integration
+
+**What:** Add `/social-strategy audit` as an optional prerequisite in the `/autoplan` pipeline, alongside `/strategist`.
+
+**Why:** Without this, the default planning path (`/autoplan`) bypasses social strategy entirely. Social context should inform scope and narrative decisions in CEO review.
+
+**Context:** `/autoplan` currently runs CEO, design, and eng reviews. Social strategy docs at `*-social-strategy-*.md` should be discoverable alongside strategy docs. Simplest integration: `/autoplan` reads social docs if they exist and mentions `/social-strategy` as a prerequisite option.
+
+**Effort:** S (human: ~1 day / CC: ~10 min)
+**Priority:** P3
+**Depends on:** Both `/social-strategy` and `/strategist` being stable
+
 ## Completed
 
 ### CI eval pipeline (v0.9.9.0)

diff --git a/VERSION b/VERSION
@@ -1 +1 @@
-0.13.3.0
+0.13.4.0
-Original file line number
+Diff line change
@@ Expand Up / @@ -17,3 +17,4 @@ extension/.auth.json @@
     .env.*
     !.env.example
     supabase/.temp/
+    SRD