chore(flags): expose flag_definition_cache_provider

[matheus-vb]

Problem

The PostHog monolith wants to use a HyperCacheFlagProvider (see posthog#50737) to read flag definitions directly from Redis instead of polling the API every 90 seconds per worker. However, the SDK's module-level API has no way to accept an external cache provider — the flag_definition_cache_provider parameter only exists on the Client constructor, not as a module-level variable.

Solution

Module-level flag_definition_cache_provider

Expose flag_definition_cache_provider as a module-level variable in posthog/__init__.py (matching the pattern used by all other SDK settings like personal_api_key, poll_interval, etc.) and pass it through setup() to the Client constructor. This allows the monolith to set posthoganalytics.flag_definition_cache_provider = HyperCacheFlagProvider(...) before any flags are evaluated.

Changes

posthog/__init__.py

• Added flag_definition_cache_provider module-level variable (defaults to None)
• Passed it through setup() to the Client constructor

How did you test this code?

• Existing tests pass (make test)

Downstream dependency

This PR must be merged and released before posthog#50737 can land, since the monolith PR sets posthoganalytics.flag_definition_cache_provider at startup.

[greptile-apps]

_{Last reviewed commit: 875b64b}

[github-actions]

posthog-python Compliance Report

Date: 2026-03-11 23:15:32 UTC
Duration: 159373ms

✅ All Tests Passed!

29/29 tests passed

---

Capture Tests

✅ 29/29 tests passed

View Details

Test	Status	Duration
Format Validation.Event Has Required Fields	✅	518ms
Format Validation.Event Has Uuid	✅	1507ms
Format Validation.Event Has Lib Properties	✅	1508ms
Format Validation.Distinct Id Is String	✅	1507ms
Format Validation.Token Is Present	✅	1507ms
Format Validation.Custom Properties Preserved	✅	1507ms
Format Validation.Event Has Timestamp	✅	1507ms
Retry Behavior.Retries On 503	✅	9519ms
Retry Behavior.Does Not Retry On 400	✅	3506ms
Retry Behavior.Does Not Retry On 401	✅	3508ms
Retry Behavior.Respects Retry After Header	✅	9515ms
Retry Behavior.Implements Backoff	✅	23518ms
Retry Behavior.Retries On 500	✅	7515ms
Retry Behavior.Retries On 502	✅	7512ms
Retry Behavior.Retries On 504	✅	7514ms
Retry Behavior.Max Retries Respected	✅	23522ms
Deduplication.Generates Unique Uuids	✅	1505ms
Deduplication.Preserves Uuid On Retry	✅	7515ms
Deduplication.Preserves Uuid And Timestamp On Retry	✅	14523ms
Deduplication.Preserves Uuid And Timestamp On Batch Retry	✅	7504ms
Deduplication.No Duplicate Events In Batch	✅	1508ms
Deduplication.Different Events Have Different Uuids	✅	1507ms
Compression.Sends Gzip When Enabled	✅	1507ms
Batch Format.Uses Proper Batch Structure	✅	1507ms
Batch Format.Flush With No Events Sends Nothing	✅	1005ms
Batch Format.Multiple Events Batched Together	✅	1506ms
Error Handling.Does Not Retry On 403	✅	3508ms
Error Handling.Does Not Retry On 413	✅	3509ms
Error Handling.Retries On 408	✅	7510ms

[haacked]

The PR description mentions additional changes (stale cache fallback for failed flags in client.py, types.py, and a new test), but I don't see those in this PR.

Is the PR description incorrect or did you forget to push?

[matheus-vb]

Oops those changes were supposed to be on another PR, will clean up the description

[haacked]