From 32de2215504ef935e6e0936bb2eec90d79ad8ff4 Mon Sep 17 00:00:00 2001 From: Krzysztof kuhy Rudnicki Date: Sun, 23 Aug 2026 16:13:28 +0200 Subject: [PATCH] Fix the Python tests CI workflow: drop mitmproxy Python tests has been red since ~2026-08-19, failing in "Install dependencies" with pip's resolution-too-deep. That is an install failure, not a test failure -- which means no Python test has actually run in CI for days, on a repo whose Python bar is 100% branch coverage. The CI log names the cause: the resolver backtracks through Brotli 1.0.6 -> 1.0.4 -> 1.0.1 and asgiref 3.8.0 -> 3.7.1 -> 3.6.0, all dragged in by mitmproxy>=10.0, then gives up. Nothing imports it: `git grep mitmproxy` finds it only in a pacman whitelist and a docs file, and check_requirements_sync.py passes without it. Reproduced and fixed locally rather than guessed, on Python 3.11.11 with pip 26.2.1 -- the versions CI uses: pip install --dry-run -r exit 1, resolution-too-deep pip install --dry-run -r exit 0 Left standing: 78 of the 80 requirements are unpinned >= ranges, so this can recur the next time a package with a wide dependency graph is added. CI_GREEN_SKIP=1: this commit IS the fix for the red baseline. Co-Authored-By: Claude Opus 5 Claude-Session: https://claude.ai/code/session_01PJrqKERAo5QUeEHWS5Asgp --- ...sts-ci-resolution-too-deep-2026-08-23.json | 36 +++++++++++++++++++ meta/requirements.txt | 1 - 2 files changed, 36 insertions(+), 1 deletion(-) create mode 100644 docs/superpowers/evidence/fix-python-tests-ci-resolution-too-deep-2026-08-23.json diff --git a/docs/superpowers/evidence/fix-python-tests-ci-resolution-too-deep-2026-08-23.json b/docs/superpowers/evidence/fix-python-tests-ci-resolution-too-deep-2026-08-23.json new file mode 100644 index 00000000..86228568 --- /dev/null +++ b/docs/superpowers/evidence/fix-python-tests-ci-resolution-too-deep-2026-08-23.json @@ -0,0 +1,36 @@ +{ + "intent": "Turn the Python tests CI workflow green. It has failed since ~2026-08-19 in 'Install dependencies' with pip's resolution-too-deep error, so no Python test has actually run in CI for days.", + "scope": [ + "meta/requirements.txt (root requirements.txt is a symlink to it)", + "Non-goal: pinning the other 78 unpinned dependencies, which would be a much larger change" + ], + "changes": [ + "Removed mitmproxy>=10.0. The CI log shows the resolver backtracking endlessly through Brotli 1.0.6/1.0.4/1.0.1 and asgiref 3.8.0/3.7.1/3.6.0, all pulled in by mitmproxy, before giving up.", + "It is safe to remove: nothing under python_pkg/ imports mitmproxy, and `git grep mitmproxy` finds it only in a pacman whitelist and a docs file. check_requirements_sync.py passes without it." + ], + "verification": [ + { + "command": "python3.11 -m venv; pip install --dry-run -r ", + "result": "fail", + "evidence": "exit 1, 'error: resolution-too-deep' -- the CI failure reproduced locally on Python 3.11.11 with pip 26.2.1, the same versions CI uses" + }, + { + "command": "pip install --dry-run -r in the same venv", + "result": "pass", + "evidence": "exit 0, resolution completes" + }, + { + "command": "python3 meta/scripts/check_requirements_sync.py --package python_pkg", + "result": "pass", + "evidence": "exit 0 -- no python_pkg import needs the dropped package" + } + ], + "risks": [ + "If a future script wants mitmproxy it must be re-added, and the resolver problem returns. Pinning it to an exact version would be the way back in.", + "The underlying fragility remains: 78 of 80 requirements are unpinned >= ranges, so another wide-dependency package could reintroduce this." + ], + "rollback": [ + "Re-add 'mitmproxy>=10.0' to meta/requirements.txt in sorted position.", + "Expect Python tests CI to go red again at the Install dependencies step." + ] +} diff --git a/meta/requirements.txt b/meta/requirements.txt index b997f2ec..20e174c8 100644 --- a/meta/requirements.txt +++ b/meta/requirements.txt @@ -46,7 +46,6 @@ lxml>=5.0 markdownify>=1.2.0 matplotlib>=3.0 mccabe>=0.7.0 -mitmproxy>=10.0 mypy>=1.8.0 numpy>=1.20 opencv-python>=4.0