Explore AI Agent Skills & Claude Prompts

process-wide singletons that bind async resources at app/worker startup leak Futures across per-test event loops under pytest-asyncio — guard the wire-up with PYTEST_CURRENT_TEST, conftest reset alone is NOT enough

schedule Updated 29 days ago

git-deploy-poller-fetch-merge-refstore-split

A git-based deploy poller that `git fetch`es in one directory but `git merge`s in another silently no-ops — and rebuilds stale code forever — when the two directories don't share a ref store. Happens when one project's working dir is a standalone clone instead of a worktree of the shared bare repo the poller fetches into.

static-ontology-knowledge-graph-trap

Hard-coded note_type / category enums in a knowledge system create filing cabinets, not knowledge graphs. The trap: classification looks like success (notes neatly distributed across folders) while the actual graph value (emergent connections, surprising links) stays at zero. Static ontology + LLM classifier = sophisticated tagger, not graph thinking.

sqlite-naive-datetime-system-tz-silent-shift

SQLite + SQLAlchemy `DateTime(timezone=True)` round-trip silently strips tzinfo. Downstream `.astimezone()` on the naive datetime then uses the SYSTEM tz, not UTC — silently shifts time-of-day classification by the local UTC offset (e.g. ~13h on a KST host, ~9h on JST, ~5h on EST). Symptom: bucket/window/cutoff classifiers (session hours, after-hours/overnight splits, daily aggregates) misclassify in production. Tests pass on the dev machine that happens to match the writer's intended tz; bug surfaces on any other host.

schedule Updated 21 days ago

local-llm-context-vs-generation-budget

Local LLMs (ollama, llama.cpp) declare huge context windows (200k+ tokens) but generation time scales with input length. On consumer GPUs, glm-4.7-flash with 16k char input times out at 300s; same model with 5k chars finishes in 50-100s. Cap derived budget for local models — declared context ≠ practical generation budget.

local-llm-agent-loop-temperature-zero-trap

Setting `temperature=0` for a tool-calling agent loop on a smaller local LLM (qwen3-coder:30b, Llama-3 30B-class, glm-4.7-flash 29.9B etc.) tightens variance BUT cuts mean accuracy hard — sometimes 5–10×. Greedy decoding locks the model into a wrong first-token path that subsequent rounds can't escape. The same sampling that creates run-to-run noise is also what gives the loop room to self-correct.

test-against-source-contracts

Test Against Source Contracts — verify tests match actual API/interface contracts

e2e-mock-shape-drift

E2E test mock fixtures using wrong API response shape — passes silently because frontend handles malformed data gracefully

json-cache-uuid-key-type-drift

When list_X comes from a JSON cache (UUID→str) but related rows come fresh from the DB (still UUID), dict[record["id"]] lookups silently miss. Symptom — first request after deploy works, all later requests behave as if related rows are empty.

canary

Post-deploy canary monitoring. Watches the live app for console errors, performance regressions, and page failures using the browse daemon. Takes periodic screenshots, compares against pre-deploy baselines, and alerts on anomalies. Use when: "monitor deploy", "canary", "post-deploy check", "watch production", "verify deploy". (gstack)

asyncio-lock-non-reentrant-deadlock

Python asyncio.Lock is NOT reentrant — adding locks to fix race conditions can introduce deadlocks when a locked method calls another locked method

shared-directory-namespace-collision

When a new feature introduces strict-naming sub-trees under an existing top-level directory that legacy code already writes to, recursive scans crash on legacy files unless the scan filters to canonical patterns first. Local unit tests pass on empty fixtures; CI fails the moment a real seeded vault/repo exists.