Skip to content

feat: release v0.8.0 high-leverage judge path - #3

Merged
iarjunganesh merged 2 commits into
mainfrom
feat/p1-high-leverage
Jul 17, 2026
Merged

iarjunganesh merged 2 commits into
mainfrom
feat/p1-high-leverage

Conversation

@iarjunganesh

Copy link
Copy Markdown
Owner

What changed

  • Adds the in-app Ask DRIFT grounded-chat experience and cross-platform judge path.
  • Makes no-key fixture claim evidence verifiable against explicit synthetic source files, with full-byte hashes, exact spans, source URLs, and raw-item relationships protected by regression coverage.
  • Ships the high-leverage README, architecture, branding, screenshot, and submission-record improvements.
  • Bumps source metadata and locked frontend/Python package roots to v0.8.0.

Why

The feature branch improves the judge-facing product while preserving DRIFT's evidence boundary. The fixture demo must be inspectable without implying a synthetic example is a real upstream release.

Impact

Operators receive API-contract-aligned Ask DRIFT validation and accessible status feedback. Fixture users can inspect source-backed examples locally with no API key. Hosted v0.7.0 remains the only deployment claimed verified until the Git-connected v0.8.0 rollout is observed.

Root cause addressed

The prior fixture claims hashed individual excerpts and reused a raw-item identifier across inconsistent source URLs. Live claims instead hash one full frozen source per raw item. This PR makes the fixture follow that same invariant and guards it with a test.

Validation

  • Ruff and mypy pass.
  • 150 tests pass with 100.00% backend coverage.
  • Clean locked frontend install, TypeScript check, and production build pass.
  • Docker Compose config, Markdown relative links, fixture byte-hash/span verification, and git diff --check pass.

P1 #5 — Add the Ask DRIFT grounded-chat box (frontend/app/AskDrift.tsx): a
single client component that posts to /chat and renders the grounded answer,
source citations, model/audit label, and grounded insight IDs, with explicit
loading, no-match (404), and error states. Contract verified against the
fixture backend; production frontend build passes.

P1 #4 — Restructure the README top for first impression: The Problem → What
DRIFT Does → Try DRIFT in 60 Seconds (hosted demo links + one-command
docker compose up judge path). Reframe the closing Disclaimer as a positive
Trust Model; keep the Codex/GPT-5.6 usage section prominent. Embed the
reviewed-briefing and claim-evidence screenshots as live-state evidence and
rename the screenshot assets to clean kebab-case filenames.

P1 #6 — Make the Quick Start cross-platform: promote docker compose up as the
one-command local path and add bash/curl equivalents beside every PowerShell
command.

Check off the P1 items in submission/NEXT_STEPS.md and record the change in
CHANGELOG (Unreleased). Video work (P0.3) is intentionally left untouched.
Fixtures: back every example claim with explicitly synthetic, checked-in source text and validate its full-byte SHA-256, exact span, URL, and raw-item relationship.

Frontend: align Ask DRIFT input validation and accessible states with the public /chat contract.

Delivery: release v0.8.0 source metadata and synchronize the README, security policy, runbook, architecture, ADRs, submission tracker, locks, and test count without claiming unverified hosted deployment.

Verification: Ruff, mypy, 150 tests at 100% backend coverage, clean frontend install/build, Docker Compose config, fixture byte-hash/span checks, Markdown link checks, and git diff --check pass.
Copilot AI review requested due to automatic review settings July 17, 2026 18:54
@vercel

vercel Bot commented Jul 17, 2026 •

Copy link
Copy Markdown

The latest updates on your projects. Learn more about Vercel for GitHub.

Project Deployment Actions Updated (UTC)
drift Ready Ready Preview, Comment Jul 17, 2026 6:55pm

@chatgpt-codex-connector

Copy link
Copy Markdown

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.

@codecov

codecov Bot commented Jul 17, 2026

Copy link
Copy Markdown

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

@iarjunganesh
iarjunganesh merged commit 8872b39 into main Jul 17, 2026
9 checks passed

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

Releases DRIFT v0.8.0 as a source release by adding an in-app grounded “Ask DRIFT” chat path, hardening fixture claim-evidence integrity against checked-in synthetic sources, and synchronizing versioning + judge-facing documentation/branding to reflect the updated trust boundary without claiming unverified hosted deployment.

Changes:

  • Add a new grounded chat UI component (AskDrift) wired to POST /chat, with loading/empty/error states and citation rendering.
  • Introduce synthetic fixture source files + a regression test that verifies fixture claim evidence spans, URLs, and SHA-256 hashes against frozen source bytes.
  • Sweep version bumps and documentation/architecture assets to v0.8.0, including updated submission/judge materials and diagrams.

Reviewed changes

Copilot reviewed 39 out of 65 changed files in this pull request and generated 4 comments.

Show a summary per file
File Description
uv.lock Bumps Python package root version to 0.8.0.
tests/unit/test_fixture_evidence.py Adds regression test ensuring fixture claim evidence matches frozen synthetic source files.
tests/integration/test_api.py Updates health/version assertion to 0.8.0.
submission/SUBMISSION.md Adds initiative pointer for the v0.8.0 source release work.
submission/NEXT_STEPS.md Updates audit/verdict notes to distinguish hosted v0.7.0 vs source v0.8.0.
submission/DRIFT_FREEZE_PLAN.md Adds a “freeze plan” document for judging-focused execution.
submission/DEMO_SCRIPT.md Minor wording update for demo script (“brand banner”).
SECURITY.md Updates supported-version table and clarifies 0.8.0 is source-only pending hosted verification.
README.md Restructures judge path, adds Ask DRIFT positioning, updates architecture assets and trust-model prose.
pyproject.toml Bumps Python project version to 0.8.0.
frontend/package.json Bumps frontend version to 0.8.0.
frontend/package-lock.json Bumps lockfile package versions to 0.8.0.
frontend/app/page.tsx Adds Ask section link and mounts the AskDrift component in the homepage flow.
frontend/app/globals.css Adds styles for Ask DRIFT form/answer UI and a visually-hidden label utility.
frontend/app/AskDrift.tsx New grounded-chat client component calling POST /chat and rendering citations + metadata.
docs/RUNBOOK.md Updates runbook to note fixture links resolve to synthetic source text.
docs/INITIATIVES.md Adds Initiative 10 entry and updates counts/coverage references.
docs/CODEX_PROMPTS.md Updates prompt checklist/status to reflect completed ops (branch protection, Codecov) and new initiative.
docs/BUILD_SEQUENCE.md Records completion of v0.8.0 source work and repo operations verification.
docs/ARCHITECTURE.md Updates architecture narrative + asset references; documents fixture evidence now points to synthetic source files.
docs/adr/README.md Updates ADR index text for ADR-010 summary alignment with v0.8.0 source release.
docs/adr/010-claim-evidence-and-review-gate.md Adds v0.8.0 fixture addendum describing synthetic evidence verification.
docs/adr/007-vercel-railway-deployment.md Terminology tweak (“hero” → “brand” banner).
CONTRIBUTING.md Updates architecture-asset source-of-truth guidance and initiative count.
CHANGELOG.md Adds 0.8.0 changelog entry covering Ask DRIFT UI, fixture evidence hardening, docs/assets updates, ops verification.
backend/fixtures/source_evidence/101.txt Adds synthetic source text for fixture raw item 101.
backend/fixtures/source_evidence/102.txt Adds synthetic source text for fixture raw item 102.
backend/fixtures/source_evidence/103.txt Adds synthetic source text for fixture raw item 103.
backend/fixtures/source_evidence/104.txt Adds synthetic source text for fixture raw item 104.
backend/fixtures/insights.json Updates fixtures to include typed claims + evidence spans/URLs/SHA-256 for inspectable claim evidence.
backend/core/config.py Bumps API version string to 0.8.0.
assets/brand/README.md Documents locked palette and banner generation workflow.
assets/brand/drift-banner-light.svg Replaces the light banner SVG output (now generator-produced).
assets/brand/build_banner.py Adds banner generator source-of-truth script.
assets/brand/build_banner_raster.mjs Adds script to rasterize banner SVGs to PNG via frontend sharp.
assets/architecture/README.md Updates architecture asset inventory and regeneration instructions for new diagram split.
assets/architecture/build_arch.py Adds hand-authored presentation architecture diagram generator.
assets/architecture/build_arch_raster.mjs Adds script to rasterize presentation architecture SVGs to PNG via frontend sharp.
assets/architecture/architecture-diagram.mmd Removes old Mermaid diagram source in favor of renamed pipeline diagram.
assets/architecture/arch-pipeline.mmd Adds renamed Mermaid pipeline diagram source with trust-zone subgraphs.
assets/architecture/arch-pipeline-light.config.json Adds Mermaid light-theme configuration for pipeline diagram rendering.
assets/architecture/arch-pipeline-dark.config.json Adds Mermaid dark-theme configuration for pipeline diagram rendering.
AGENTS.md Updates current-phase status to reflect v0.8.0 as source release and refreshed verification notes.
.gitattributes Preserves fixture source-evidence .txt bytes across platforms for stable SHA-256 verification.
Files not reviewed (1)
  • frontend/package-lock.json: Generated file

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Comment thread frontend/app/AskDrift.tsx
Comment on lines +30 to +36
async function ask(pending: string) {
const trimmed = pending.trim();
if (trimmed.length < MIN_QUESTION_LENGTH || status === "loading") return;

setStatus("loading");
setResult(null);
setErrorMessage("");
Comment thread frontend/app/AskDrift.tsx
Comment on lines +149 to +153
{result.source_citations.map((citation) => (
<a key={citation} href={citation} target="_blank" rel="noreferrer">
Source ↗
</a>
))}
Comment on lines +4 to +8
import { readFileSync, writeFileSync } from "node:fs";
import { createRequire } from "node:module";

const require = createRequire("file://" + process.cwd() + "/../../frontend/");
const sharp = require("sharp");
Comment on lines +3 to +7
import { readFileSync, writeFileSync } from "node:fs";
import { createRequire } from "node:module";

const require = createRequire("file://" + process.cwd() + "/../../frontend/");
const sharp = require("sharp");
@iarjunganesh
iarjunganesh deleted the feat/p1-high-leverage branch July 17, 2026 22:40

This branch was successfully deployed

1 active deployment
Preview — 962c57a6 Deployed Jul 17, 2026 by vercel[bot]
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants