diff --git a/tests/e2e/AGENTS.md b/tests/e2e/AGENTS.md
index 94035cfe849..b6abcdb6ba2 100644
--- a/tests/e2e/AGENTS.md
+++ b/tests/e2e/AGENTS.md
@@ -98,7 +98,7 @@ Each suite provides its own `client` fixture (see `llm_translation/passthrough_c
 
 Request and response bodies are typed pydantic models in `models.py`; only the fields a test reads are modelled, and nothing passes raw dicts. Outcomes come back as a `Result[R]` tagged union (`Success`, `NetworkError`, `UnauthorizedError`, `RateLimitedError`, `ValidationError`, `UnknownApiError`). Handle them with `match`, or call `unwrap(...)` when a non-success should fail the test. The harness hard-fails and never skips: a test marked `e2e` fails when no proxy answers its liveness probe, and once a request reaches the proxy any wrong behavior is likewise a hard failure, so a missing proxy turns the run red instead of being mistaken for a pass
 
-Mark live tests with `@pytest.mark.e2e` (on the class or the module). Add `@pytest.mark.quiet_stack` to a test that measures the proxy itself (RSS, latency): the shared stack lock in `stack_lock.py` then runs it while no other test on the host is hitting the stack, marked or not, so the reading depends only on the test's own traffic. Use `scoped_key` for a fresh all-models key that auto-deletes, `resources` when you need to create and tear down more than a key, and `unique_marker()` from `e2e_config` to keep prompts, tags, and customer ids from colliding across concurrent runs and the shared response cache
+Mark live tests with `@pytest.mark.e2e` (on the class or the module). Coverage of the harness itself carries no marker and runs whether or not a proxy is up. Add `@pytest.mark.quiet_stack` to a test that measures the proxy itself (RSS, latency): the shared stack lock in `stack_lock.py` then runs it while no other test on the host is hitting the stack, marked or not, so the reading depends only on the test's own traffic. Use `scoped_key` for a fresh all-models key that auto-deletes, `resources` when you need to create and tear down more than a key, and `unique_marker()` from `e2e_config` to keep prompts, tags, and customer ids from colliding across concurrent runs and the shared response cache
 
 ## Record and replay fixtures
 
@@ -251,7 +251,7 @@ other.<area>.<case>.<assertion>
 ```
 
 ## Hard Rules
-- no unit tests of any kind under `tests/e2e`. a product feature is proven end to end against a live proxy, never with a unit test, and the harness itself is not unit-tested here either. no monkeypatching or mock tests. if a contributor asks you to write an end to end test, do NOT stage a unit test with it; if you find a product gap, call it out in the PR description
+- no unit tests of a product feature under `tests/e2e`, and no mock tests or monkeypatching of code anywhere in it: a product feature is proven end to end against a live proxy, never with a unit test. if a contributor asks you to write an end to end test, do NOT stage a unit test with it; if you find a product gap, call it out in the PR description. the harness's own plumbing is the one exception: the markerless tests in the root-level `test_*.py` files, `coverage_registry/test_collector.py`, `guardrails/test_guardrails_client.py`, the `claude_code/_*_unit_tests/` trees, and the `load/` aggregation tests carry no `e2e` marker, run without a proxy, and take their inputs as arguments or env vars (setting an env var through pytest's `monkeypatch` fixture is fine, patching a function, class, or module is not), and no coverage-registry or compat-matrix cell rests on them. judge a change inside one of them by that standard, not as a misplaced product test
 
 - use model management endpoints to create new models for a test. this could be in a conftest / inline for each test. ask the user what they want.