[blogger-rewire] repoint routine paths to claude-knowcap hub + seed persona rotation

[Smetools]

What

Repoints the blogger routine spec at the claude-knowcap hub and makes persona rotation resolvable. The routine's source docs (brand DNA, personas, SEO/competitor research, screenshots) migrated out of this repo to the hub in #40, but the spec still pointed at the now-deleted docs/brand/ + docs/research/ paths — so a live /knowcap-blogger run could not resolve a persona, keyword audit, competitor doc, or screenshot index.

Changes

• routines/blogger/ROUTINE.md + routines/_skills/write-blog-draft/SKILL.md — every brand/research/screenshot path repointed to ../claude-knowcap/knowledge/... (sibling hub repo, relative to repo root). Added a "Knowledge base (hub) — paths" table as the single source of truth.
• Persona rotation — documented state.json shape ({personas[], cursor}) and made it robust to a missing file: a fresh clone / cloud routine defaults to index 0 (odoo-partners) and creates it, so a run always resolves a persona. The persona set is the research-gated ICP roster; the cursor is a blind round-robin queue.
• docs/content-pipeline/drafts/ — recreated (removed in the migration) as the routine's draft output target.
• .gitignore — ignore routines/**/state.json (runtime cursor, not tracked).
• Screenshot caveat — hub screenshots aren't web-served; must be copied into the web app's public assets before embedding (a hub-relative path would 404 on the live site). Moot until the screenshot index is populated (currently empty).

Verification

All 6 hub paths confirmed resolving from repo root (PRODUCT-PERSONAS.md, VISION.md, POSITIONING.md, SEO-AUDIT-*.md, competitors/read.ai/positioning.md, screenshots/_index.json). state.json confirmed gitignored. .vercel/ excluded from the commit.

Canonical /knowcap-blogger skill note (in AV/skills/, separate repo) updated in parallel to reflect the sibling-hub path model.

🤖 Generated with Claude Code

Summary by CodeRabbit

• Chores

  • Updated .gitignore to exclude routine state files from version control.
  • Added directory placeholder for blog draft content pipeline.

• Documentation

  • Updated blog routine specifications to reference knowledge sources from sibling repository instead of local paths.
  • Documented updated blog routine selection logic for content types and persona rotation.

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
knowcap-landing	Ready	Preview, Comment	Jun 14, 2026 7:59am

[coderabbitai]

Caution

Review failed

Pull request was closed or merged during review

📝 Walkthrough

Walkthrough

The PR migrates the blogger routine and write-blog-draft skill specs from local docs/brand/ and docs/research/ knowledge paths to the sibling ../claude-knowcap/ hub repository. It also rewrites the mode-selection algorithm in ROUTINE.md, adds a persona round-robin state spec, inserts a web-serving caveat for screenshots, and adds .gitignore and .gitkeep scaffolding.

Changes

Hub-based blog routine migration

Layer / File(s)	Summary
Repo scaffolding .gitignore, docs/content-pipeline/drafts/.gitkeep	Adds routines/**/state.json to .gitignore and creates the docs/content-pipeline/drafts/ directory placeholder.
SKILL.md hub path redirects and web-serving caveat routines/_skills/write-blog-draft/SKILL.md	Redirects vision_md, positioning_md, persona_section_md, competitor_positioning_md, and available_screenshots[].file inputs to ../claude-knowcap/ hub paths. Adds a caveat that hub screenshot paths must be copied to app/public/blog/<slug>/ before embedding.
ROUTINE.md hub paths, persona rotation, and mode-selection algorithm routines/blogger/ROUTINE.md	Rewrites knowledge-base section to reference ../claude-knowcap/ for all brand/competitor/persona assets. Expands state.json spec with JSON shape and cursor-advancement rules. Replaces high-level mode block with an explicit case-study → comparison → thesis selection algorithm. Realigns the Inputs section to new hub paths.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Poem

🐰 Hoppity hop to the hub we go,
brand DNA now lives in sibling repo's glow.
Round-robin personas, modes one, two, three—
screenshots need copying (from 404 set free!).
State files stay hidden, drafts find their place,
the blogger routine runs with newly mapped grace. ✨

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and concisely summarizes the main changes: repointing routine paths to the claude-knowcap hub and implementing persona rotation. It directly reflects the primary objectives of the PR.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch hassan-blogger-rewire

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[Smetools]

Added commit 7261fcc (gap-1 fix from the dry-run): routines/blogger/keyword-opportunities.md — per-persona keyword table as the routine's PRIMARY keyword source (input #5), since the only hub SEO audit is technical/on-page with no keyword table. Google Trends demoted to advisory/tiebreaker (negligible MENA volume for B2B ICP — closest odoo term ≤0.9/100). Also flags the roster-vs-research mismatch: mena-agencies is in the rotation roster but the segmentation study rates it an anti-buyer.