open/linkedin-studio
1
0
Fork 0
Code Issues Releases 1 Activity
246 commits 1 branch 6 tags 11 MiB
Commit graph

7 commits

Author SHA1 Message Date
Kjell Tore Guttormsen
e17134ee9b refactor(linkedin-studio): S31c descriptive-%-scrub — platform-norm percentages asserted as fact -> SSOT 
24 edits / 12 files (+26/-26). Unsourced platform/algorithm/audience percentages reconciled to
SSOT vocabulary (figure/proportion/multiplier unverified). Catalog + new sibling clusters
(64% follow-up x5, wrong-window 70% x4, Stage-2 6-10% x2) + borderlines (70% retention, 70%
mobile) + the ~3% save-worthy straggler (surfaced, not silent). The SSOT-sourced ~70% reach
figure is KEPT; only the wrong window corrected (60min/1h -> first 15-30 min). Sourced/computed
benchmarks kept (Buffer 178%/247%, Socialinsider 11%). KEPT C1: ~45% AI-comment figure (already
hedged correlational/medium-confidence). Gate 81/0/0 exit 0, counts 29/19/26 + v0.5.0.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016qgzo6rxthw7KuxHjn5vyE
 2026-06-20 09:54:44 +02:00
Kjell Tore Guttormsen
1bab1d6df1 refactor(linkedin-studio): S30 magnitude-scrub (discrete-% class) — unsourced reach/engagement penalties -> SSOT 
Hardening-class, NOT re-hardening: surgical SSOT-reconciliation of discrete percentage
penalties/declines stated as fact with no primary source in the SSOT
(references/algorithm-signals-reference.md). Same tool-grounded discipline as S27/S28
(read-and-show -> grep-confirm -> re-grep final). Re-grep surfaced drift + same-class siblings
beyond the plan's stored list; all surfaced and operator-approved before edit.

Scope: the discrete-% reach/engagement-penalty class only. The unsourced "Nx" reach/format
MULTIPLIER class (~50 instances across ~15 files) is a separate, larger pass -> deferred to S31
(operator: run everything, across multiple sessions).

HARDEN (20 edits, 7 files):
- linkedin-formats.md (5): :6 47-50% decline + :7 15%->31% feed-share -> directional; :176
  AI-comment 30%/55% -> ~45% less engagement (correlational, medium); :231/:279 hashtags -68%
  -> diminishing returns, no discrete figure.
- linkedin-growth-playbook (6): :158 47-50% decline (twin of formats:6) + :166 hashtags -68%
  + :224/:225 post-length 25%/32% + :435/:828 posting-frequency 25% -> directional, no
  discrete figure. (:221 1.17x multiplier folded in per operator approval; the rest of the
  multiplier class -> S31.)
- glossary.md (2): :91 engagement-bait -30-50% + :235 topic-gap -15-25% -> "correlate with
  lower reach, no discrete figure".
- engagement-coach.md (2): :195 55% + :455 -30%/-55% AI-comment -> ~45% less engagement
  (correlational), actively suppressed.
- post-feedback-monitor.md (1): :330 -25%/post -> "tends to split your own audience".
- ab-testing-framework.md (1): :66 hashtags -68% -> no discrete figure.
- poll-strategy-guide.md (2): :20 / :205 poll-overuse penalty -> declining effectiveness
  (directional).

KEPT INTACT (operator-locked / different class): engagement-pod + AI-slop suppression framing
(SSOT high-confidence); firsthour:112 (no number); poll:206 / poll:3 (qualitative); growth:567
conversion-rate; formats:268 list item.

VERIFY: discrete-% penalty/decline class re-grep across the 7 files -> NONE; leave-items intact;
bash scripts/test-runner.sh -> Passed 81 / Failed 0 / Warnings 0, exit 0; counts 29/19/26/6 +
v0.5.0 unchanged (.md prose only). Disposition: FIXED (20 edits, 7 files), one atomic commit,
local (push held).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016qgzo6rxthw7KuxHjn5vyE
 2026-06-20 09:08:59 +02:00
Kjell Tore Guttormsen
6bd263144f docs(linkedin-studio): M0-15 — repoint remaining prose families + route ab-tests/plans 2026-06-18 13:04:18 +02:00
Kjell Tore Guttormsen
ad5bb702f6 fix(linkedin-studio): promote post-feedback-monitor to Opus (Opus-default) 
post-feedback-monitor was the lone non-Opus/Sonnet agent (model: haiku). It
does human-facing, real-time post-publish coaching — exactly the workload the
standing Opus-default targets — so the Haiku setting both contradicts the
default and underpowers the surface the new /linkedin:firsthour command hands
off to (remediation Step 18).

- agents/post-feedback-monitor.md: model haiku -> opus.
- CLAUDE.md: agent-table model column Haiku -> Opus. No Haiku agent remains.

Verify: grep 'model: haiku' agents/ -> none; CLAUDE.md row shows Opus;
structure lint -> 62/62, counts unchanged.

Plan Step 18 (Wave 4 S3). Counts unchanged.
[skip-docs]: tre-doc + version bump deferred to Step 21 per remediation plan.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
 2026-05-30 05:36:30 +02:00
Kjell Tore Guttormsen
3dd8f40c31 fix(linkedin-studio): propagate reconciled algorithm numbers, cite-not-restate 
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
 2026-05-29 20:32:17 +02:00
Kjell Tore Guttormsen
52fa87a255 refactor(linkedin)!: rename plugin linkedin-thought-leadership → linkedin-studio (v3.0.0) 
BREAKING CHANGE: the marketplace slug, the agent namespace
(linkedin-studio:<agent>), and the runtime state-file path
(~/.claude/linkedin-studio.local.md) all change. Reinstall required;
existing state migrated in place (post metrics, streak, history preserved).
The /linkedin:* commands are unchanged — the command namespace is set
per-command in frontmatter and was always independent of the plugin slug.
Functionality is byte-identical to v2.4.0; this release is pure identity.

- dir + manifests: plugins/linkedin-studio + plugin.json + root marketplace.json
- agent namespace updated in commands/newsletter.md (only functional invoker)
- state path updated in 4 hook scripts + topic-rotation prompt + state template
- catch-all skill dir renamed skills/linkedin-studio (5 functional skills unchanged)
- docs + version bump to 3.0.0 across README badge, CHANGELOG, root README/CLAUDE.md
- historical records (CHANGELOG past entries, docs/ build artifacts,
  config-audit v5.0.0 snapshots) intentionally retain the old slug

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
 2026-05-29 11:32:02 +02:00
Kjell Tore Guttormsen
40986575b6 feat(ultraplan-local): v1.6.0 — /ultraresearch-local deep research command 
Add /ultraresearch-local for structured research combining local codebase
analysis with external knowledge via parallel agent swarms. Produces research
briefs with triangulation, confidence ratings, and source quality assessment.

New command: /ultraresearch-local with modes --quick, --local, --external, --fg.
New agents: research-orchestrator (opus), docs-researcher, community-researcher,
security-researcher, contrarian-researcher, gemini-bridge (all sonnet).
New template: research-brief-template.md.

Integration: --research flag in /ultraplan-local accepts pre-built research
briefs (up to 3), enriches the interview and exploration phases. Planning
orchestrator cross-references brief findings during synthesis.

Design principle: Context Engineering — right information to right agent at
right time. Research briefs are structured artifacts in the pipeline:
ultraresearch → brief → ultraplan --research → plan → ultraexecute.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
 2026-04-08 08:58:35 +02:00