[Testing 21] test depth: mutation testing + SSE load harness (on-demand) - #254
Open
amal66 wants to merge 2 commits into
Open
[Testing 21] test depth: mutation testing + SSE load harness (on-demand)#254amal66 wants to merge 2 commits into
amal66 wants to merge 2 commits into
Conversation
|
|
amal66
force-pushed
the
olp-pr/test-depth-stretch
branch
from
July 25, 2026 21:31
daece4a to
8ff496e
Compare
amal66
force-pushed
the
olp-pr/test-depth-stretch
branch
from
July 26, 2026 14:32
8ff496e to
549832f
Compare
amal66
force-pushed
the
olp-pr/test-depth-stretch
branch
from
August 3, 2026 01:57
549832f to
84dfa56
Compare
Collaborator
|
Please rebase to main and then split the PRs into two separate ones. Thanks. |
Two on-demand depth tools, neither a merge gate: - Stryker mutation testing scoped to the security-critical backend libs (access.ts, downloadTokens.ts, safeError.ts, chat/citations.ts). Measured 74.0-76.4% mutation score across runs; thresholds.break=69 fails only on real regressions. `npm run test:mutation` locally (~3 min), .github/workflows/mutation.yml on demand + monthly cron, HTML report uploaded as artifact. - k6 load harness for the SSE chat stream (loadtest/sse-stream.js): ramps to N concurrent POST /chat streams, checks TTFB and that every stream delivers events through to the [DONE] sentinel — the past incident class (streams timing out on long tool calls). Lenient, documented thresholds. .github/workflows/loadtest.yml is workflow_dispatch-only and boots nothing: point it at a staging stack (PR Open-Legal-Products#210). docs/test-depth.md explains how to read the mutation report, how to run the k6 harness against the local stack, and why neither gates merges. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…s with self-contained target guidance WHY THIS MATTERS The load-test harness (workflow, doc, and k6 script) all told operators to point it at "the staging stack from PR Open-Legal-Products#210". PR Open-Legal-Products#210 was CLOSED unmerged — it was rejected for leaking admin credentials — so the one piece of infrastructure the harness claims to require does not exist. A newcomer following the docs would hit a dead end: the reference reads as "there is a blessed staging stack somewhere", when in reality there is no such stack and never will be from that PR. Worse, linking to a credential-leaking PR as recommended reading is itself a small security smell. Docs that point at dead artifacts erode trust in all the other docs around them. WHAT IS A SELF-CONTAINED REFERENCE Documentation can either point at an artifact ("use the stack from PR Open-Legal-Products#210") or describe a contract ("use any stack that serves X behind Y"). Pointers are fragile: PRs get closed, branches get deleted, staging environments get torn down, and the doc silently rots. A contract-style reference survives all of that because it tells the operator what the target must PROVIDE rather than where one specific instance lived. For an on-demand load harness — which by design boots nothing itself — the contract is the only stable thing to document. HOW THE FIX WORKS Every "Open-Legal-Products#210" reference in the three harness files is replaced with the actual contract the target stack must satisfy: - a deployed, NON-production backend the operator owns, - serving the backend API's streaming endpoint: POST {BASE_URL}/chat Authorization: Bearer <supabase access token> - with real LLM provider keys configured (each iteration performs a real chat completion). Concretely: - .github/workflows/loadtest.yml: the header comment and the `target_url` input description now describe that contract, e.g. description: "Backend base URL of a non-production stack you deployed (serving the backend API with auth), ..." - docs/test-depth.md: the "Running from GitHub Actions" section describes the same contract instead of naming PR Open-Legal-Products#210. - loadtest/sse-stream.js: the header comment does likewise. No behavior changes — the workflow inputs, k6 scenario, thresholds, and safety warnings ("never point at production") are untouched. Verified by grepping the three files for "Open-Legal-Products#210" (no matches), YAML-parsing the workflow, and `node --check` on the k6 script. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
amal66
force-pushed
the
olp-pr/test-depth-stretch
branch
from
August 6, 2026 03:36
349b538 to
64ddcf4
Compare
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
3 participants
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.
Two tools that go a level deeper than the regular test suite. Neither one gates merges — both are run-on-demand, so they cost you nothing day to day.
1. Mutation testing for the security-critical code. Line coverage says a test ran a line; it doesn't say a test would notice if that line's behavior changed. Mutation testing checks exactly that: Stryker plants hundreds of small deliberate bugs (flip a comparison, delete an early return, weaken a regex) and re-runs the suite for each one. Any bug the suite doesn't catch is listed as "survived" — a concrete missing assertion. It's scoped to the four places where a hollow test would be dangerous: sharing/access checks (
access.ts), signed download tokens (downloadTokens.ts), secret redaction (safeError.ts), and citation extraction (chat/citations.ts). Current score: ~74–76% of planted bugs caught; the run fails only if that drops below 69, so it flags real regressions, not noise. Run it withcd backend && npm run test:mutation(~3 min) or from the Actions tab; a monthly scheduled run catches slow drift. The HTML report shows each surviving mutant inline in the code — the survivingsafeErrorredaction-regex mutants are ready-made test cases for whoever wants a small follow-up PR.2. A load test for chat streaming. The SSE stream behind
POST /chatis the hot path and has caused incidents before (streams timing out during long tool calls).loadtest/sse-stream.js(k6) ramps up to N simultaneous chat streams and verifies each one starts quickly and actually runs to completion ([DONE]arrives) — with deliberately lenient, documented thresholds: a red run means "streams are hanging," not "we missed an SLO we never set." Run it locally perdocs/test-depth.md, or via the manual SSE load test workflow, which boots nothing itself — you point it at an already-running staging stack (the one from #210 is the intended target) and store a test user's token in theLOADTEST_AUTH_TOKENsecret. Never point it at production: it creates real chats and burns real tokens.Why not merge gates? Mutation testing multiplies test runtime; the load test needs a live stack with real keys. Both are too slow/stateful to put in front of every PR, and a flaky required check is worse than none. If the project gains contributors and a permanent staging stack, the natural upgrade path is in
docs/test-depth.md.Verified: two full local Stryker runs completed green (2m51s / 2m25s); k6 script validated via
k6 inspect+ a dry run in docker; both workflows YAML-parse. Lockfile changes are npm-generated (npm install, never hand-edited).🤖 Generated with Claude Code