v1.1.0

[1.1.0] — 2026-04-18

Context Efficiency Release. SHINE 1.0 auto-enabled all 16 plugins on install, pushing baseline context to ~95k — combined with user-configured MCPs and a large CLAUDE.md, many sessions crossed Claude Code's effective token budget and became unusable. 1.1 introduces context profiles, slims the always-loaded instruction file, and makes plugin activation opt-in and switchable.

Added

• Context profile system (shine/profiles/*.json) — 6 curated bundles declaring enabledPlugins + disabledMcpjsonServers + estimatedContextTokens:
  • minimal (~15k) — no plugins. New default.
  • writing (~20k) — context7.
  • outbound (~35k) — context7 + prospecting MCPs kept (Apollo, Gmail, CommonRoom).
  • seo (~40k) — context7 + Ahrefs kept.
  • dev (~70k) — full engineering stack (serena, LSPs, playwright, supabase, code-simplifier, superpowers).
  • full (~95k) — all 16 plugins (parity with previous default).
• shine-profile.cjs — Node CLI that atomically rewrites ~/.claude/settings.json's enabledPlugins and disabledMcpjsonServers from a selected profile. Commands: list, show <name>, activate <name>, current. Atomic via fs.renameSync from .tmp.
• shine bash shim (shine/bin/shine) — unified CLI dispatching to shine-profile.cjs (activate|list|show|current|profile) and shine-tools.cjs (doctor|onboard|help|index-skills|--version). Add ~/.claude/shine/bin to PATH to call shine directly.
• install.sh --profile=<name> — CLI flag and interactive picker (6 options, default minimal). Installer now parses the selected profile JSON via node to determine which plugin@marketplace entries to install — instead of the fixed 16-plugin loop.
• shine/references/integrations-map.md — full task-to-integration routing (was CLAUDE.md §§1-14). Loaded on demand.
• shine/references/mcp-capability-map.md — BYO MCP inventory (was CLAUDE.md §15).
• shine/references/agency-playbook.md — tone switching, GDPR guard, proposal assembly, lead enrichment (was CLAUDE.md §§17-20).
• shine/references/tool-tiers.md — free-first fallback rule (was CLAUDE.md §21).
• shine/references/decision-rules.md — full 29-rule matrix (was CLAUDE.md Decision Rules).

Changed

• settings.template.json — the key fix. enabledPlugins now defaults to {} (was 16 entries = true), alwaysThinkingEnabled flipped to false, disabledMcpjsonServers added with explanatory comment. Baseline context drops from ~95k to ~15k.
• CLAUDE.md — split from 31,798 bytes to ~6,300 bytes (80% reduction). Slim core retains identity, auto-use principle, top-level decision rules, and the full §16 Factual / RAG Discipline table (unchanged, MANDATORY). Detailed routing, MCP maps, agency playbooks, and the 29-rule matrix moved to shine/references/*.md for on-demand loading.
• install.sh — profile-aware install flow. Copies shine-profile.cjs, the shine shim, and shine/profiles/*.json; runs shine-profile activate at end of install to sync settings.json with the chosen profile. Help text and next-steps output document the shine activate command.
• shine/VERSION — bumped 1.0.3 → 1.1.0.
• README.md — new "Context profiles" subsection under Quick install, --profile flag documented in flags table, plugin section now describes profile-gated install.
• QUICKSTART.md — install step renumbered and extended with profile picker, new step 3 "Switch context profile", troubleshooting rows for heavy-context symptoms and missing MCPs.

Fixed

• hooks/shine-check-update.js — corrected hardcoded repo slug from the stale diShine-digital-agency/shine-claude-code-framework to diShine-digital-agency/SHINE-Code-System. Previously every update-check silently 404'd.

Migration from 1.0.x

1. Pull the latest SHINE-Code-System and re-run ./install.sh — your existing ~/.claude/ is backed up atomically before any change.
2. Pick a profile when prompted (suggestion: minimal for marketing work, dev for coding, outbound for prospecting).
3. Switch any time with shine activate <name> and restart Claude Code.
4. No data loss: your memory/, MEMORY.md, and custom skills/agents are preserved.

Why this matters

The SHINE 1.0 default combined with a typical user's claude.ai MCP server set easily crossed 200k tokens of context before the first prompt. On Haiku 4.5 sessions with a 200k window, this meant repeated compaction, missing tool schemas, and degraded answer quality. The 1.1 profile system restores headroom: a minimal SHINE install + 5 user MCPs now sits comfortably under 40k, leaving 160k+ for actual work.

v1.0.0

[1.0.0] — 2026-04-16

Added

• shine-tools.cjs index-skills — scans ~/.claude/skills/, parses every SKILL.md frontmatter, groups by inferred category (core / agency / marketing / sales / consulting / tech / ops / data / brand / misc) and writes skills/INDEX.md (140 entries across 10 categories).
• shine-tools.cjs onboard — non-destructive first-run wizard. Creates 4 starter memory files (preference-communication.md, preference-stack.md, preference-factual-rag.md, style-email-it.md) plus MEMORY.md index. Skips any file that already exists.
• shine-tools.cjs doctor — prints a JSON health report: ~/.claude/ presence, critical file checks, counts of agents / skills / hooks / memories. Useful for CI and support triage.
• SHINE_HOOK_FORMAT=json — when set, every SHINE hook emits a single JSONL line per notification instead of plain stderr text. Enables ingestion into log pipelines without breaking the default human-readable mode. Applies to shine-context-monitor, shine-prompt-guard, shine-read-guard, shine-check-update, shine-learning-log.
• hooks/_shine-hook-output.js — shared emitter used by all stderr-emitting hooks. Single source of truth for text / JSON formatting, fail-open semantics, and level tagging (info / warn / block).
• hooks/shine-learning-log.js — new Stop-event hook. Appends one JSONL line per Claude turn to ~/.claude/memory/learning-log.jsonl with ts, cwd, last_tool, transcript_bytes, event. No conversation content is ever captured — log is PII-free by design. LRU-trims to SHINE_LEARNING_LOG_MAX (default 10000 lines). Controls: SHINE_DISABLE_LEARNING_LOG=1.
• skills/shine-retro/SKILL.md — reads the learning log, aggregates tool usage / cwd coverage / long-session counts, and proposes concrete memory updates. Write-safe: only creates external-retro-<date>.md, never edits existing memory files.
• skills/shine-pipeline/SKILL.md — compose skills into an ad-hoc sequential workflow with a shared markdown scratchpad. Hard 8-step cap, destructive-step confirmation, context-budget awareness.
• skills/shine-tour/SKILL.md — interactive 6-minute guided tour for first-time users. Runs 14 install-health checks with copy-paste fixes, teaches the 5 SHINE pillars (rules / memory / skills / agents / hooks), demos a real read-only workflow, writes an audit trail to ~/.claude/sessions/tour-<ts>.md. Flags: --section <name>, --check-only, --fix. Includes a 20-row troubleshooting matrix covering every known failure mode (missing claude CLI, bad settings.json, hook path drift, MCP auth, false-positive prompt-guard, log bloat, Windows/WSL, perms, backup cleanup).
• shine/templates/watermark.md — verified-source watermark template. Inline labels (_[verified — …]_, _[unverified — pattern inferred]_, _[drafted — no source]_) + mandatory Sources footer for every client-facing deliverable with factual claims.

Changed

• CLAUDE.md §16 (Factual / RAG Discipline) — adds the watermark convention as the enforceable output format for mixed deliverables.
• shine/references/ui-brand.md — new §10b "Verified-source watermark" pointing to the template.
• settings.template.json — registers shine-learning-log.js on the Stop event.
• install.sh — log message bumped to reflect 141 skills, 8 lifecycle hooks (adds Stop), 6 templates (adds watermark).

Verified

• node shine/bin/shine-tools.cjs index-skills → writes skills/INDEX.md with 140 entries.
• node shine/bin/shine-tools.cjs onboard → creates 5 files on first run, skips on second run.
• node shine/bin/shine-tools.cjs doctor → returns a valid JSON health report.
• SHINE_HOOK_FORMAT=json regression: text mode unchanged, JSON mode emits parseable JSONL, exit codes preserved (prompt-guard still exits 2 on block).
• .github/scripts/check_agent_quality.py → all agents still pass (no regressions).

---

[0.2.0]

Added

• Tier-1 workflows fleshed out — do, fast, ship, next, resume-project, map-codebase, review, complete-milestone, new-milestone, manager, stats, session-report now contain real step-by-step logic, decision gates, error handling, and output formats.
• Tier-2 workflows fleshed out — add-phase, insert-phase, remove-phase, add-tests, check-todos, cleanup, audit-milestone, audit-uat, analyze-dependencies, forensics, docs-update, secure-phase, validate-phase, ui-phase, ui-review now have concrete protocols.
• shine/bin/shine-tools.test.cjs — zero-dep test suite with 16 tests covering version, help, slug (ASCII + Unicode), frontmatter (inline, block scalars, missing), state (--set with/without value, values containing =), commit (no-message error), roadmap (parse + empty), workstream list, unknown command.
• .github/scripts/check_references.py — validates every @~/.claude/... reference resolves to a real file; now gated in CI.
• QUICKSTART.md — 1-page onboarding (vs 30KB README).
• shine-tools edge-case hardening:
  • --version / -v flag prints JSON from shine/VERSION.
  • state --set without value errors clearly instead of silently no-op'ing.
  • parseFrontmatter now handles YAML block scalars (| literal, > folded).
• Token-aware context monitoring — hooks/shine-context-monitor.js estimates tokens (~bytes/3.5) and warns on soft 120k / hard 180k token thresholds in addition to byte thresholds.
• Output templates added to 8 previously thin agency agents: shine-pm-coordinator (Asana/Linear queries + digest template), shine-crm-operator (CRM query patterns + dedup confidence scoring), shine-lead-scorer (4-axis scoring model), shine-martech-architect (4-layer stack design + mermaid), shine-content-editor (FK readability + inline-edit template), shine-translator (glossary format + back-translation drift check), shine-brand-voice-auditor (4-axis tone rubric + consistency formula), shine-sales-strategist (pipeline + forecast template).

Changed

• install.sh — now also copies shine-tools.test.cjs and runs it post-install (non-blocking).
• .github/workflows/validate.yml — adds check_references.py + shine-tools.test.cjs as CI gates.

Verified

• node shine/bin/shine-tools.test.cjs → 16 passed, 0 failed.
• python3 .github/scripts/check_references.py → 116 references, all resolve.
• node shine/bin/shine-tools.cjs --version → {"version":"1.0.2","name":"shine-tools"}.

---

[0.1.1]

Added

• shine/bin/shine-tools.cjs — 18-subcommand Node.js runtime engine (no external deps, safe spawnSync only). Subcommands: init, state, commit, commit-to-subrepo, roadmap, verify, frontmatter, resolve-model, generate-slug, websearch, config-get, config-set-model-profile, phase, requirements, summary-extract, history-digest, in, workstream, help. Unblocks all 21 engineering agents and 18+ skills that depended on it.
• shine/workflows/*.md — 53 workflow files. 10 fully-fleshed (Tier 1: execute-phase, plan-phase, discuss-phase, verify-work, new-project, quick, autonomous, help, health, progress); 43 structured stubs for the rest following the same pattern.
• shine/templates/*.md — 5 scaffold templates (project.md, requirements.md, context.md, milestone-archive.md, UAT.md).
• shine/references/*.md — 2 reference files (ui-brand.md covering voice, formatting, MoSCoW, commit messages; questioning.md covering 5-Why, MoSCoW prompts, discovery-call template, red flags).
• install.sh — now copies shine/bin/, shine/workflows/, shine/templates/, shine/references/, validates 9 critical files post-install, and smoke-tests the runtime.

Fixed

• Duplicate skill removed: skills/sync/ (kept skills/shine-sync/).
• All 61 missing files identified in the audit now exist, tracked, and validated by CI.

Verified

• python3 .github/scripts/check_frontmatter.py agents → 39 agents OK.
• python3 .github/scripts/check_frontmatter.py skills → 138 skills OK.
• python3 .github/scripts/check_hook_refs.py → 7 hooks, all resolve.
• python3 .github/scripts/check_secrets.py → no secret-shaped strings.
• node shine/bin/shine-tools.cjs help → runtime responds.

---

[0.1.0]

First release of the SHINE framework under the diShine digital agency.

Added

• 39 agents under agents/ — 21 core engineering (shine-planner, shine-executor, shine-verifier, shine-debugger, shine-security-auditor, shine-codebase-mapper, shine-roadmapper, shine-ui-auditor, shine-ui-checker, shine-ui-researcher, shine-integration-checker, shine-nyquist-auditor, shine-plan-checker, shine-phase-researcher, shine-project-researcher, shine-research-synthesizer, shine-advisor-researcher, shine-assumptions-analyzer, shine-user-profiler, shine-doc-verifier, shine-doc-writer) + 18 agency (shine-client-researcher, shine-brand-voice-auditor, shine-competitor-scout, shine-proposal-writer, shine-sales-strategist, shine-content-editor, shine-seo-strategist, shine-martech-architect, shine-data-analyst, shine-copywriter, shine-gdpr-analyst, shine-translator, shine-pm-coordinator, shine-account-manager, shine-crm-operator, shine-retro-facilitator, shine-lead-scorer, shine-persona-researcher).
• 138 skills under skills/ across 8 categories:
  • 61 core SHINE skills
  • 17 early agency skills (proposal, draft-email, gdpr-audit, lead-enrich, client-brief, kickoff, retrospective, roi-estimate, seo-audit, tag-audit, cookie-scan, pii-safe, compliance-ai, meta-check, status-report, sync, client-tone)
  • 10 marketing/content (content-calendar, blog-post, social-post, newsletter, `l...