Misc reports

What a Team seat actually buys vs an individual Pro/Max plan: usage multipliers, Claude Code + Claude Tag availability, admin/collaboration features, and the flat-vs-metered billing flip. For if LIFE turns on a Team workspace — verified against the live pricing docs.

A plain-English manual for the Transcriber app: sign in, upload a recording, and get back a clean transcript with the speakers separated, saved to the Knowledge Base. Covers the status labels, supported files, privacy, tips, and troubleshooting. Written for non-technical users.

Audit of keyless (Workload Identity Federation) secrets and OpenTofu across the LIFE stack — a keep/retire verdict with bucketed recommendations.

Full 30-day audit mapping Claude Code work to Jira issues — the detailed version behind the leadership view.

Executive summary of the 30-day Claude-Code-to-Jira audit — what shipped, framed for stakeholders.

Audit of the golden-prompt judge rubric — stat chips, charts, and top-edit findings across judges.

Vendor research ranking transcription and diarization options, with tradeoffs and a recommended stack.

Audit of Data Lab export formats and training-data readiness — severity-graded findings, a pipeline diagram, and a letter-grade verdict.

Retrospective on a Data Lab quality pass, with stat cards and verdict scoring.

Classifies Knowledge Base documents for promotion (move / transform / leave / dup / hold) across corpora, with table-of-contents navigation.

Tracks the Wisdom prompt's benchmark score across versions through v0.8.62 (weighted, unweighted, and safety series).

What we're building, why it can work, and exactly where we are. The bet (predict each user's reaction, then personalise around it), the evidence base (psychometric ceilings, Meta/TikTok behavioural modelling, Centaur), persona-collapse risk, our leak-free eval, and an honest status read — measuring instrument proven on synthetic data, real-user readout gated on three unblocks. Prepared for the CEO & CFO.

Four platform initiatives shipped (Unified Prompt Registry, Hub Identity Authority, KB deployment, Internal Admin Panel) and a full Data Lab UX overhaul (steering, per-candidate judging, voice playback + six supporting improvements). 420 commits across 9 repos. Visual report with screenshots of every hub-linked app.

The 16-criterion contemplative-alignment rubric. Affective, self-construct, and cognitive dimensions, plus a meta skilful-means test. Browse by criterion.

Turning reviewer hours into a training-data asset. Pipeline, current state, and outstanding issues.

Side-by-side comparison of Wisdom system-prompt versions across the golden-question set (10 scenarios).

Stratified 35-scenario gate (all safety levels, variant B + C) confirming the v0.8.3 regression.

Executive overview of the LIFE company KB: architecture, content model, and why we're building it.

Walkthrough of the wiki-ingest skill: source pages, canonical topics, patch behavior, review gates.

Strategic plan from the 2026-04 roadmap session. Sprints, dependencies, and what we're betting on.

Historical snapshot. LIFE-bot supersedes this Slack bot; do not use the old endpoints as current deploy targets.

Progress report on the Golden Prompt judge: rubric iteration, agreement metrics, and per-scenario scores.

Weekly status for the LIFE app across product, alignment, and infra.