Spec - #20
Merged
Merged
Spec#20
Conversation
…nd' for service management
- Updated test for pricing to ensure no prefix fallback for unknown variants. - Added normalization tests for model versioning with dot and dash formats. - Adjusted compact advise logic to require higher utilization percentage for compaction. - Introduced new tests for model routing recommendations based on task and prior errors. - Simplified installation script tests and ensured correct paths for MCP binaries. - Added smoke tests for MCP stdio protocol to validate JSON-RPC message handling. - Enhanced compact output tests to include new conditions for auto-compaction. - Improved context compressor tests to preserve recent turns and validate handover packet rendering.
- Updated tests to replace direct command invocations with `tools call` for search, read, edit, and memory operations. - Adjusted assertions to reflect changes in output format and structure. - Modified expected tool names in tests to align with new context-based architecture. - Enhanced session optimization tests to include new savings reporting and routing features. - Added new tests for session compaction and model routing savings. - Ensured all tests reflect the transition from task-based to context-based tool handling.
- New package src/atelier/infra/benchmarks/ with publisher.py, markdown_renderer.py
- Jinja2 template templates/post.md.j2 for weekly benchmark posts
- PublishReport dataclass aggregates routing, compact, quality and replay results
- Delta comparison vs prior week (reports/YYYY-W(N-1)/benchmark.json)
- atelier benchmark publish CLI command (--since, --output, --corpus, --dry-run)
- Reports written to reports/YYYY-Www/benchmark.{md,json} + reports/index.json
- 30 passing tests covering all acceptance criteria
- First real publication committed in reports/2026-W20/
Closes spec 05 (Shipped)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Backend (9 new FastAPI routes):
- GET /v1/sessions, /v1/sessions/{id}
- GET /v1/memory/facts, /v1/memory/facts/{fact_id}
- GET /v1/insights, /v1/outcomes/summary, /v1/outcomes/{session_id}
- GET /v1/reports, /v1/reports/{week}
- 26 passing unit tests in tests/core/service/test_api_week2_routes.py
Frontend TypeScript client (frontend/src/api.ts):
- 10 new interfaces: SessionSummary, SessionReport, TopTool, MemoryFact,
InsightsSessionSummary, OutcomesSummary, Opportunity, InsightsWindow,
ReportMeta, ReportContent
- 9 new API methods
New pages:
- Sessions.tsx — sortable table of recent sessions, click → detail
- SessionDetail.tsx — full SessionReport (cost breakdown, top tools, savings)
- Outcomes.tsx — route/compact outcome scores with empty state
- Reports.tsx — benchmark report list + react-markdown renderer
- Telemetry.tsx — old Insights content moved here
Updated pages:
- Insights.tsx — rebuilt to show InsightsWindow (top sessions, vendor/tool cost, opportunities)
- Memory.tsx — added cross-vendor tab as default; knowledge blocks as second tab
- App.tsx — new routes (/sessions/:id, /traces, /memory, /insights, /telemetry, /outcomes, /reports); updated NAV_ITEMS
Tests:
- Sessions.test.tsx, SessionDetail.test.tsx, Outcomes.test.tsx, Reports.test.tsx (new)
- Insights.test.tsx, Memory.test.tsx (updated)
- All 27 new/updated tests passing; Watchdogs pre-existing failure unchanged
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
…ndbox and actionable savings - Restructured Optimizations page for a cleaner, linear flow (Summary -> Configuration -> Comparison -> History -> Actions -> Evidence). - Added interactive Policy Sandbox with sliders and switches for tuning quality floor, context trigger, and compaction features. - Updated candidate grid to display potential savings in green rather than absolute cost. - Removed redundant CandidateSummaryCard and Task mix sections. - Refactored backend `_pct_change` to return fractional values for consistent UI display. - Adjusted `Economy` and `Maximum Saving` presets to use distinct compaction triggers and correct quality penalties.
- Removed unused imports in various test files to clean up the codebase. - Updated test names for clarity and consistency, such as changing `test_context_reuse_inject_runtime_reasoning` to `test_context_reuse_inject_context`. - Enhanced error handling in tests by skipping tests when required fixtures or conditions are not met. - Adjusted assertions in tests to reflect changes in expected outputs and improved test coverage for new features. - Renamed and refactored functions related to monitoring and watchdogs for better alignment with their functionality. - Consolidated test logic for better maintainability and readability, particularly in tests related to session parsing and artifact handling.
pankaj4u4m
added a commit
that referenced
this pull request
Jun 18, 2026
pankaj4u4m
added a commit
that referenced
this pull request
Jun 21, 2026
pankaj4u4m
added a commit
that referenced
this pull request
Jun 21, 2026
pankaj4u4m
added a commit
that referenced
this pull request
Jun 22, 2026
pankaj4u4m
added a commit
that referenced
this pull request
Jul 2, 2026
pankaj4u4m
added a commit
that referenced
this pull request
Jul 2, 2026
pankaj4u4m
added a commit
that referenced
this pull request
Jul 3, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.