ci: dogfood the FULL fleet — 9 cloud + 2 Macs, matching mort
Build & push image / build-and-push (pull_request) Successful in 5s

The dogfood workflow had a truncated 5-model list (3 cloud + the 2 Macs)
and was missing GADFLY_PROVIDER_LENS_CONCURRENCY. Restore mort's full
fleet so gadfly reviews its own PRs with the same 11 reviewers and the
model-quality scoreboard is comparable across both repos:

  9 cloud: minimax-m3, glm-5.2, glm-5.1, kimi-k2.7-code, deepseek-v4-pro,
           nemotron-3-super, gpt-oss:120b, qwen3-coder:480b, gemma4
  2 local: m1/qwen3:14b, m5/qwen3.6:35b-mlx

GADFLY_MODELS / *_CONCURRENCY / *_LENS_CONCURRENCY now match mort's
adversarial-review.yml verbatim.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
2026-06-27 14:31:07 -04:00
parent 1cdda32dbc
commit 48af34f4ca
+8 -4
View File
@@ -4,7 +4,7 @@
# caches :latest, and this build is what carries foreman provider-type support) # caches :latest, and this build is what carries foreman provider-type support)
# as a specialist swarm and posts # as a specialist swarm and posts
# ONE consolidated review comment as gitea-actions. Advisory only — never blocks a # ONE consolidated review comment as gitea-actions. Advisory only — never blocks a
# merge. Gadfly reviewing its OWN PRs — dogfooding, full fleet (3 cloud + the M1/M5 # merge. Gadfly reviewing its OWN PRs — dogfooding, full fleet (9 cloud + the M1/M5
# Macs), copied from mort's setup. # Macs), copied from mort's setup.
name: Adversarial Review (Gadfly) name: Adversarial Review (Gadfly)
@@ -41,7 +41,7 @@ jobs:
|| github.actor == 'fizi' || github.actor == 'fizi'
|| github.actor == 'dazed')) || github.actor == 'dazed'))
runs-on: ubuntu-latest runs-on: ubuntu-latest
# Full fleet (3 cloud + 2 local Macs, all running concurrently) reviewing # Full fleet (9 cloud + 2 local Macs, all running concurrently) reviewing
# every PR with the 3-lens suite — the slow local lanes dominate wall time. # every PR with the 3-lens suite — the slow local lanes dominate wall time.
timeout-minutes: 90 timeout-minutes: 90
steps: steps:
@@ -66,10 +66,14 @@ jobs:
# still post. (Gitea secrets aren't auto-exposed — map each explicitly.) # still post. (Gitea secrets aren't auto-exposed — map each explicitly.)
GADFLY_ENDPOINT_M1: ${{ secrets.GADFLY_ENDPOINT_M1 }} GADFLY_ENDPOINT_M1: ${{ secrets.GADFLY_ENDPOINT_M1 }}
GADFLY_ENDPOINT_M5: ${{ secrets.GADFLY_ENDPOINT_M5 }} GADFLY_ENDPOINT_M5: ${{ secrets.GADFLY_ENDPOINT_M5 }}
# 3 cloud (parallel) + M1 Pro + M5 Max — one consolidated comment each. # Full fleet: 9 cloud (3 at a time) + M1 Pro + M5 Max — one consolidated
GADFLY_MODELS: "minimax-m3:cloud,deepseek-v4-flash:cloud,glm-5.2:cloud,m1/qwen3:14b,m5/qwen3.6:35b-mlx" # comment each. Matches mort's setup so the model-quality scoreboard is
# comparable across both repos.
GADFLY_MODELS: "minimax-m3:cloud,glm-5.2:cloud,glm-5.1:cloud,kimi-k2.7-code:cloud,deepseek-v4-pro:cloud,nemotron-3-super:cloud,gpt-oss:120b-cloud,qwen3-coder:480b-cloud,gemma4:cloud,m1/qwen3:14b,m5/qwen3.6:35b-mlx"
# cloud runs 3 at once; each Mac one at a time; all three lanes parallel. # cloud runs 3 at once; each Mac one at a time; all three lanes parallel.
GADFLY_PROVIDER_CONCURRENCY: "ollama-cloud=3,m1=1,m5=1" GADFLY_PROVIDER_CONCURRENCY: "ollama-cloud=3,m1=1,m5=1"
# 3 cloud models x 3 lenses = 9 concurrent ollama-cloud queries (under the 10 budget).
GADFLY_PROVIDER_LENS_CONCURRENCY: "ollama-cloud=3"
# Default => the 3-lens suite (security, correctness, error-handling). # Default => the 3-lens suite (security, correctness, error-handling).
# Set the repo var GADFLY_SPECIALISTS to override (csv / "all" / "auto"). # Set the repo var GADFLY_SPECIALISTS to override (csv / "all" / "auto").
GADFLY_SPECIALISTS: ${{ vars.GADFLY_SPECIALISTS || 'security,correctness,error-handling' }} GADFLY_SPECIALISTS: ${{ vars.GADFLY_SPECIALISTS || 'security,correctness,error-handling' }}