feat(spike): T-008 — validate Gemini API real metrics, resolve OI-001#75
Closed
renatobardi wants to merge 1 commit intomainfrom
Closed
feat(spike): T-008 — validate Gemini API real metrics, resolve OI-001#75renatobardi wants to merge 1 commit intomainfrom
renatobardi wants to merge 1 commit intomainfrom
Conversation
- Run SP-01 spike with real GEMINI_API_KEY (ensi project, 2026-04-18) - gemini-2.5-flash: 10/10 BriefArtifact compliance, p50=6361ms p95=7980ms - gemini-2.5-pro: 5/5 PrdArtifact compliance, p50=42754ms p95=50285ms - NFR PASS: typical pipeline ~125s, worst-case p95 ~165s (within 5min limit) - Discovered: gemini-2.0-flash deprecated for new accounts → updated to gemini-2.5-flash - Discovered: maxOutputTokens must be 8192 (4096 truncates Pro-stage responses) - responseSchema overrides contradictory prompts — retry loop not triggered in 15 runs - Update architecture-artifact.json: real latency, model names, OI-001 resolved - Update sp-01-result.md: verdict GO (was GO WITH CONSTRAINTS from dry run)
|
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
GEMINI_API_KEY(ensi project, 2026-04-18)Real metrics (replacing SP-01 dry-run estimates)
gemini-2.5-flashgemini-2.5-proNFR: Typical pipeline ~125s total, worst-case p95 ~165s — well within 5-min limit.
Key findings
gemini-2.0-flashis deprecated for new accounts (404 on first run) → updated togemini-2.5-flashmaxOutputTokensmust be 8192 — 4096 truncates Pro-stage responses at ~192 linesresponseSchemaoverrides contradictory prompts — retry loop not triggered in 15 runs (safety net only)Files changed
_bmad/spikes/sp-01/src/main.rs— dual-model runner (flash for Brief, pro for PRD)_bmad/spikes/sp-01/src/gemini.rs— maxOutputTokens 4096 → 8192_bmad/spikes/sp-01/results/runs.json— real 16-entry dataset_bmad/spikes/sp-01-result.md— verdict updated GO (was GO WITH CONSTRAINTS)_bmad/docs/architecture-artifact.json— model names, real latency, OI-001 resolved, ADR-002/003 updatedAcceptance criteria coverage
results/runs.jsonarchitecture-artifact.jsonupdated