Move --font-display and --font-body CSS variable declarations from
workers/page.tsx into a new evals/layout.tsx so all evals sub-pages
(methodology, workers, workers/[roleId], etc.) inherit the font
variables without duplicating the setup.
* fix: race conditions in subtask delegation system
Comprehensive fix for race conditions and error handling gaps in the subtask
delegation system. Addresses multiple failure modes that could leave parent
tasks permanently stuck in 'delegated' status, causing nested subtasks to hang.
Key fixes:
- Remove initialStatus from taskMetadata rebuild (eliminates status overwrites)
- Persist delegation metadata to per-task files (resolves globalState eviction)
- Add delegationInProgress mutex guard (prevents concurrent delegation ops)
- TOCTOU race fixes with fresh re-reads before writes
- Abort-aware pWaitFor predicate (prevents false 60s timeout on user input)
- Remove silent .catch(() => {}) — all errors now logged unless task is aborting
- Single-attempt delegation with parent repair on failure (no retry band-aids)
- Cancel debouncedEmitTokenUsage in dispose() (prevents zombie callbacks)
- new_task isolation truncation for parallel tool calls
* fix: write all 6 delegation fields in every saveDelegationMeta call site
* fix: align delegation tests with single-attempt implementation (no retry)
- Add "Value Map: Salary vs Interview Score" scatter to comparison page
- Dots colored by tier, sized by success rate
- Sweet Spot quadrant highlight (upper-left)
- Respects existing provider/success-rate filters
- Add "AI Coding Capability Over Time" scatter to landing page
- 10 models from Jun 2025 to Feb 2026
- Dots colored by provider, sized by cost efficiency
- Dashed trend line showing upward trajectory
- Add MODEL_TIMELINE data to mock-recommendations.ts
- Atmospheric header with role-colored blur gradients
- Glass-morphism containers for chart, filters, and export
- Styled language toggle pills with role color accents
- Themed provider checkboxes and success rate slider
- Custom chart tooltip with backdrop blur
- Export buttons with press feedback
- framer-motion scroll-triggered animations
- Bottom navigation with pill-style links
- Role themes: reviewer (violet) and autonomous (cyan) added to candidates page
- Add 5 engineer roles: Junior, Senior, Staff, Architecture Reviewer, Autonomous Agent
- Build role selection landing page with hiring metaphor at /evals/workers
- Build candidate rankings page with tiered recommendations at /evals/workers/[roleId]
- Build candidate comparison page with Recharts charts at /evals/workers/[roleId]/compare
- Build "How We Interview" methodology page at /evals/methodology
- Add mock data with real eval scores from 27 model runs
- Implement "Hire This Engineer" CTA linking to Roo Code Cloud
- Implement "Configure Extension" CTA with clipboard copy
- Per-language score breakdowns (Go, Java, JS, Python, Rust)
- Daily salary pricing (80 tasks/agent/day estimate)
- framer-motion animations, glass-morphism design, role color themes
- Tone-of-voice compliance (no em dashes, no hype, workflow-first copy)
- vscode:// deep link design doc at plans/vscode-deep-link-design.md
