[eli-r-ph]

💡 Motivation and Context

v1 is missing main's 7.83.2-7.87.1: emails from code, the MCP spec alignment, $experiment_exposure (#1154), flags, errors and AI fixes, #1209 and #1211, and the Markdown reference tooling. This repo only squash-merges, and squashed syncs leave the merge base with main at October, so the 8.0 release PR (#1016) keeps conflicting (86 files).

Warning

Do not squash or merge this PR with the button. After approval it lands by fast-forwarding v1 to this exact head, which keeps both merge commits. Squashing drops them, and #1016 keeps conflicting.

Changes:

• 994adb7c is a real merge of main up to the commit before feat(flags): send $experiment_exposure for v2 experiment splits #1154. v1 already had main through 7.83.1 from the squashed chore: sync main into v1 (7.83.0–7.83.1, flags v4) #1175, so every file is merged against that point, not the old base. The commit message lists each resolution.
  • references/ ends equal to main's, so the HogRef files chore(references): keep HogRef JSON only for versions people install #1225 removed stay removed.
  • Not carried: fix(capture): retry capture v1 requests that get 404 or 429 #1178 (7.x opt-in retries; v1 has had them since fix: retry capture requests that get 404 or 429 #1132).
• c9185bda merges feat(flags): send $experiment_exposure for v2 experiment splits #1154 and the rest, as its own commit because it changes the enqueue pipeline.
  • The pinned flag attribution is applied in one place, at the end of _fill_event_defaults, which both clients use. main applies it after its super property merge; v1 fills defaults in layers instead.
  • capture_event.py carries that change, and it is the only file in this PR that main did not change.
• 268c613b regenerates the 8.x and migration references with --major 8.
• 80563b2f points main's paired benchmarks (ci: add paired Python SDK performance benchmarks #1134, ci: integrate shared Python benchmark reports #1135) at capture v1. They drove 7.x's /batch sender with gzip and capture_mode, which 8.0 removed, so every benchmark job failed here and would on main after release.
• Tests main added are adapted to v1's API (capture_ai, v1's senders, 8.0's super property rule); the commit messages list each one.

💚 How did you test it?

• Every file that differs from v1 was changed on main since the last sync, apart from capture_event.py above.
• ruff, mypy and the public API check pass; the snapshot changes only by main's email API.
• 7,378 tests pass. The 3 fork tests in test_metrics_autocapture.py fail locally on pristine v1 too (macOS fork).
• Removing the pin fails the sync and async attribution tests.
• Merging this head into current main has 0 conflicts.

📝 Checklist

• All commits are signed (see Commit signing)
• I reviewed the submitted code.
• I added tests to verify the changes.
• I updated the docs if needed.
• No breaking change or entry added to the changelog.

If releasing new changes

• Ran sampo add to generate a changeset file

No changeset: main's changes shipped in 7.83.2-7.87.1, and 8.0's changesets are on v1.

🤖 Agent context

Autonomy: Human-driven (agent-assisted)

Claude Code (Claude Opus 5.5).

🤖 Generated with Claude Code

[github-actions]

SDK performance: posthog-python

1 change worth reviewing: 1 observed regression.

Flags require at least a 5% change in the median and all six pairs agreeing on direction. This is a noise filter, not proof of a performance change or a CI gate.

Test	Measurement	Before PR	With PR	Change
Capture and deliver events (32-byte payload, burst)	Time per operation	80.98 µs/op	86.26 µs/op	6.5% more time

View all results

Test	Measurement	Before PR	With PR	Change	Assessment
Capture and deliver events (32-byte payload, burst) Shard: capture	Managed heap after settling	—	—	—	unavailable; base:​ unsupported; head:​ unsupported; base:​ CPython has no uninstrumented total managed-heap byte counter; tracing would distort this RSS pass.​; head:​ CPython has no uninstrumented total managed-heap byte counter; tracing would distort this RSS pass.​
Capture and deliver events (4,096-byte payload, burst) Shard: capture	Managed heap after settling	—	—	—	unavailable; base:​ unsupported; head:​ unsupported; base:​ CPython has no uninstrumented total managed-heap byte counter; tracing would distort this RSS pass.​; head:​ CPython has no uninstrumented total managed-heap byte counter; tracing would distort this RSS pass.​
Capture and deliver events (32-byte payload, steady) Shard: capture	Managed heap after settling	—	—	—	unavailable; base:​ unsupported; head:​ unsupported; base:​ CPython has no uninstrumented total managed-heap byte counter; tracing would distort this RSS pass.​; head:​ CPython has no uninstrumented total managed-heap byte counter; tracing would distort this RSS pass.​
Capture and deliver events (4,096-byte payload, steady) Shard: capture	Managed heap after settling	—	—	—	unavailable; base:​ unsupported; head:​ unsupported; base:​ CPython has no uninstrumented total managed-heap byte counter; tracing would distort this RSS pass.​; head:​ CPython has no uninstrumented total managed-heap byte counter; tracing would distort this RSS pass.​
Queue an event (32-byte payload) Shard: capture	Allocated memory per operation	—	—	—	unavailable; base:​ unsupported; head:​ unsupported; base:​ Background preparation has no public completion boundary before flush; see capture.​drain for full-pipeline allocations.​; head:​ Background preparation has no public completion boundary before flush; see capture.​drain for full-pipeline allocations.​
Queue an event (32-byte payload) Shard: capture	Allocations per operation	—	—	—	unavailable; base:​ unsupported; head:​ unsupported; base:​ Background preparation has no public completion boundary before flush; see capture.​drain for full-pipeline allocations.​; head:​ Background preparation has no public completion boundary before flush; see capture.​drain for full-pipeline allocations.​
Queue an event (4,096-byte payload) Shard: capture	Allocated memory per operation	—	—	—	unavailable; base:​ unsupported; head:​ unsupported; base:​ Background preparation has no public completion boundary before flush; see capture.​drain for full-pipeline allocations.​; head:​ Background preparation has no public completion boundary before flush; see capture.​drain for full-pipeline allocations.​
Queue an event (4,096-byte payload) Shard: capture	Allocations per operation	—	—	—	unavailable; base:​ unsupported; head:​ unsupported; base:​ Background preparation has no public completion boundary before flush; see capture.​drain for full-pipeline allocations.​; head:​ Background preparation has no public completion boundary before flush; see capture.​drain for full-pipeline allocations.​
Evaluate 1,000 flags under concurrent load Shard: flags-memory	Managed heap after settling	—	—	—	unavailable; base:​ unsupported; head:​ unsupported; base:​ CPython has no uninstrumented total managed-heap byte counter; tracing would distort this RSS pass.​; head:​ CPython has no uninstrumented total managed-heap byte counter; tracing would distort this RSS pass.​
Refresh 1,000 flag definitions Shard: flags-memory	Managed heap after settling	—	—	—	unavailable; base:​ unsupported; head:​ unsupported; base:​ CPython has no uninstrumented total managed-heap byte counter; tracing would distort this RSS pass.​; head:​ CPython has no uninstrumented total managed-heap byte counter; tracing would distort this RSS pass.​
Capture and deliver events (32-byte payload, burst) Shard: capture	Allocated memory per operation	131.85 KiB/op	133.2 KiB/op	1% more memory	Not flagged
Capture and deliver events (32-byte payload, burst) Shard: capture	Allocations per operation	1,325.53 count/op	1,337.53 count/op	0.9% more	Not flagged
Capture and deliver events (32-byte payload, burst) Shard: capture	Time per operation	80.98 µs/op	86.26 µs/op	6.5% more time	Worth reviewing: observed regression
Capture and deliver events (32-byte payload, burst) Shard: capture	Idle process memory (RSS)	37.55 MiB	37.61 MiB	0.1% more memory	Not flagged
Capture and deliver events (32-byte payload, burst) Shard: capture	Observed peak process memory (RSS)	40 MiB	40.04 MiB	<0.1% more memory	Not flagged
Capture and deliver events (32-byte payload, burst) Shard: capture	Process memory after settling (RSS)	40 MiB	40.04 MiB	<0.1% more memory	Not flagged
Capture and deliver events (32-byte payload, burst) Shard: capture	Process CPU time per operation	80.89 µs/op	86.82 µs/op	7.3% more time	Not flagged
Capture and deliver events (32-byte payload, burst) Shard: capture	Lifetime peak process memory (RSS)	40.2 MiB	40.22 MiB	<0.1% more memory	Not flagged
Capture and deliver events (4,096-byte payload, burst) Shard: capture	Allocated memory per operation	183.93 KiB/op	185.28 KiB/op	0.7% more memory	Not flagged
Capture and deliver events (4,096-byte payload, burst) Shard: capture	Allocations per operation	1,317.52 count/op	1,329.52 count/op	0.9% more	Not flagged
Capture and deliver events (4,096-byte payload, burst) Shard: capture	Time per operation	115.13 µs/op	111.93 µs/op	2.8% less time	Not flagged
Capture and deliver events (4,096-byte payload, burst) Shard: capture	Idle process memory (RSS)	41.03 MiB	41.29 MiB	0.6% more memory	Not flagged
Capture and deliver events (4,096-byte payload, burst) Shard: capture	Observed peak process memory (RSS)	46.81 MiB	47.04 MiB	0.5% more memory	Not flagged
Capture and deliver events (4,096-byte payload, burst) Shard: capture	Process memory after settling (RSS)	46.81 MiB	47.04 MiB	0.5% more memory	Not flagged
Capture and deliver events (4,096-byte payload, burst) Shard: capture	Process CPU time per operation	109.07 µs/op	104.72 µs/op	4% less time	Not flagged
Capture and deliver events (4,096-byte payload, burst) Shard: capture	Lifetime peak process memory (RSS)	51.3 MiB	51.51 MiB	0.4% more memory	Not flagged
Capture and deliver events (32-byte payload, steady) Shard: capture	Allocated memory per operation	136.92 KiB/op	138.27 KiB/op	1% more memory	Not flagged
Capture and deliver events (32-byte payload, steady) Shard: capture	Allocations per operation	1,370.44 count/op	1,382.42 count/op	0.9% more	Not flagged
Capture and deliver events (32-byte payload, steady) Shard: capture	Time per operation	96.89 µs/op	100.23 µs/op	3.4% more time	Not flagged
Capture and deliver events (32-byte payload, steady) Shard: capture	Idle process memory (RSS)	37.39 MiB	37.57 MiB	0.5% more memory	Not flagged
Capture and deliver events (32-byte payload, steady) Shard: capture	Observed peak process memory (RSS)	37.58 MiB	37.73 MiB	0.4% more memory	Not flagged
Capture and deliver events (32-byte payload, steady) Shard: capture	Process memory after settling (RSS)	37.58 MiB	37.73 MiB	0.4% more memory	Not flagged
Capture and deliver events (32-byte payload, steady) Shard: capture	Process CPU time per operation	95.81 µs/op	99.59 µs/op	3.9% more time	Not flagged
Capture and deliver events (32-byte payload, steady) Shard: capture	Lifetime peak process memory (RSS)	39.46 MiB	39.46 MiB	No observed change	Not flagged
Capture and deliver events (4,096-byte payload, steady) Shard: capture	Allocated memory per operation	189.1 KiB/op	190.45 KiB/op	0.7% more memory	Not flagged
Capture and deliver events (4,096-byte payload, steady) Shard: capture	Allocations per operation	1,362.37 count/op	1,374.37 count/op	0.9% more	Not flagged
Capture and deliver events (4,096-byte payload, steady) Shard: capture	Time per operation	113.45 µs/op	113.59 µs/op	0.1% more time	Not flagged
Capture and deliver events (4,096-byte payload, steady) Shard: capture	Idle process memory (RSS)	40.83 MiB	41.01 MiB	0.4% more memory	Not flagged
Capture and deliver events (4,096-byte payload, steady) Shard: capture	Observed peak process memory (RSS)	41.33 MiB	41.48 MiB	0.3% more memory	Not flagged
Capture and deliver events (4,096-byte payload, steady) Shard: capture	Process memory after settling (RSS)	40.91 MiB	41.06 MiB	0.4% more memory	Not flagged
Capture and deliver events (4,096-byte payload, steady) Shard: capture	Process CPU time per operation	107.58 µs/op	107.39 µs/op	0.2% less time	Not flagged
Capture and deliver events (4,096-byte payload, steady) Shard: capture	Lifetime peak process memory (RSS)	41.23 MiB	41.37 MiB	0.3% more memory	Not flagged
Queue an event (32-byte payload) Shard: capture	Caller-thread CPU time per operation	59.79 µs/op	58.56 µs/op	2.1% less time	Not flagged
Queue an event (32-byte payload) Shard: capture	Time per operation	61.02 µs/op	58.55 µs/op	4% less time	Not flagged
Queue an event (4,096-byte payload) Shard: capture	Caller-thread CPU time per operation	62.62 µs/op	66.82 µs/op	6.7% more time	Not flagged
Queue an event (4,096-byte payload) Shard: capture	Time per operation	62.62 µs/op	66.82 µs/op	6.7% more time	Not flagged
Evaluate all 10 flags Shard: flags-scaling	Allocated memory per operation	198.17 KiB/op	200.98 KiB/op	1.4% more memory	Not flagged
Evaluate all 10 flags Shard: flags-scaling	Allocations per operation	2,276.33 count/op	2,339.33 count/op	2.8% more	Not flagged
Evaluate all 10 flags Shard: flags-scaling	Time per operation	88.79 µs/op	88.16 µs/op	0.7% less time	Not flagged
Evaluate all 10 flags Shard: flags-scaling	Process CPU time per operation	88.79 µs/op	88.16 µs/op	0.7% less time	Not flagged
Evaluate all 100 flags Shard: flags-scaling	Allocated memory per operation	1.9 MiB/op	1.93 MiB/op	1.4% more memory	Not flagged
Evaluate all 100 flags Shard: flags-scaling	Allocations per operation	22,383.56 count/op	23,013.56 count/op	2.8% more	Not flagged
Evaluate all 100 flags Shard: flags-scaling	Time per operation	851.38 µs/op	858.97 µs/op	0.9% more time	Not flagged
Evaluate all 100 flags Shard: flags-scaling	Process CPU time per operation	851.41 µs/op	858.64 µs/op	0.8% more time	Not flagged
Evaluate all 1,000 flags Shard: flags-scaling	Allocated memory per operation	18.97 MiB/op	19.24 MiB/op	1.4% more memory	Not flagged
Evaluate all 1,000 flags Shard: flags-scaling	Allocations per operation	223,354.94 count/op	229,654.94 count/op	2.8% more	Not flagged
Evaluate all 1,000 flags Shard: flags-scaling	Time per operation	8.6 ms/op	8.54 ms/op	0.6% less time	Not flagged
Evaluate all 1,000 flags Shard: flags-scaling	Process CPU time per operation	8.6 ms/op	8.54 ms/op	0.6% less time	Not flagged
Evaluate 1,000 flags under concurrent load Shard: flags-memory	Idle process memory (RSS)	43.47 MiB	43.07 MiB	0.9% less memory	Not flagged
Evaluate 1,000 flags under concurrent load Shard: flags-memory	Observed peak process memory (RSS)	43.75 MiB	43.38 MiB	0.9% less memory	Not flagged
Evaluate 1,000 flags under concurrent load Shard: flags-memory	Process memory after settling (RSS)	43.75 MiB	43.38 MiB	0.9% less memory	Not flagged
Evaluate 1,000 flags under concurrent load Shard: flags-memory	Lifetime peak process memory (RSS)	43.7 MiB	43.88 MiB	0.4% more memory	Not flagged
Refresh 1,000 flag definitions Shard: flags-memory	Idle process memory (RSS)	43.12 MiB	43.06 MiB	0.1% less memory	Not flagged
Refresh 1,000 flag definitions Shard: flags-memory	Observed peak process memory (RSS)	43.12 MiB	43.07 MiB	0.1% less memory	Not flagged
Refresh 1,000 flag definitions Shard: flags-memory	Process memory after settling (RSS)	43.12 MiB	43.07 MiB	0.1% less memory	Not flagged
Refresh 1,000 flag definitions Shard: flags-memory	Lifetime peak process memory (RSS)	42.98 MiB	42.86 MiB	0.3% less memory	Not flagged
Evaluate one boolean flag Shard: flags-single	Allocated memory per operation	14.74 KiB/op	15.06 KiB/op	2.2% more memory	Not flagged
Evaluate one boolean flag Shard: flags-single	Allocations per operation	166.76 count/op	173.76 count/op	4.2% more	Not flagged
Evaluate one boolean flag Shard: flags-single	Time per operation	9.81 µs/op	9.75 µs/op	0.6% less time	Not flagged
Evaluate one boolean flag Shard: flags-single	Process CPU time per operation	9.81 µs/op	9.75 µs/op	0.6% less time	Not flagged
Evaluate a cohort-filtered flag Shard: flags-single	Allocated memory per operation	22.81 KiB/op	23.14 KiB/op	1.4% more memory	Not flagged
Evaluate a cohort-filtered flag Shard: flags-single	Allocations per operation	263.27 count/op	270.27 count/op	2.7% more	Not flagged
Evaluate a cohort-filtered flag Shard: flags-single	Time per operation	12.98 µs/op	13.04 µs/op	0.4% more time	Not flagged
Evaluate a cohort-filtered flag Shard: flags-single	Process CPU time per operation	12.98 µs/op	13.04 µs/op	0.4% more time	Not flagged
Evaluate a flag (early match) Shard: flags-single	Allocated memory per operation	35.4 KiB/op	35.73 KiB/op	0.9% more memory	Not flagged
Evaluate a flag (early match) Shard: flags-single	Allocations per operation	380.27 count/op	387.27 count/op	1.8% more	Not flagged
Evaluate a flag (early match) Shard: flags-single	Time per operation	18.49 µs/op	18.52 µs/op	0.2% more time	Not flagged
Evaluate a flag (early match) Shard: flags-single	Process CPU time per operation	18.49 µs/op	18.52 µs/op	0.2% more time	Not flagged
Evaluate a flag (late match) Shard: flags-single	Allocated memory per operation	53.28 KiB/op	53.61 KiB/op	0.6% more memory	Not flagged
Evaluate a flag (late match) Shard: flags-single	Allocations per operation	569.27 count/op	576.27 count/op	1.2% more	Not flagged
Evaluate a flag (late match) Shard: flags-single	Time per operation	26.2 µs/op	26.39 µs/op	0.7% more time	Not flagged
Evaluate a flag (late match) Shard: flags-single	Process CPU time per operation	26.2 µs/op	26.39 µs/op	0.7% more time	Not flagged
Evaluate a flag (no match) Shard: flags-single	Allocated memory per operation	51.83 KiB/op	52.16 KiB/op	0.6% more memory	Not flagged
Evaluate a flag (no match) Shard: flags-single	Allocations per operation	549.77 count/op	556.77 count/op	1.3% more	Not flagged
Evaluate a flag (no match) Shard: flags-single	Time per operation	24.91 µs/op	25.27 µs/op	1.5% more time	Not flagged
Evaluate a flag (no match) Shard: flags-single	Process CPU time per operation	24.91 µs/op	25.27 µs/op	1.5% more time	Not flagged
Evaluate a property-filtered flag Shard: flags-single	Allocated memory per operation	17.52 KiB/op	17.85 KiB/op	1.9% more memory	Not flagged
Evaluate a property-filtered flag Shard: flags-single	Allocations per operation	191.27 count/op	198.27 count/op	3.7% more	Not flagged
Evaluate a property-filtered flag Shard: flags-single	Time per operation	10.68 µs/op	10.72 µs/op	0.4% more time	Not flagged
Evaluate a property-filtered flag Shard: flags-single	Process CPU time per operation	10.68 µs/op	10.72 µs/op	0.4% more time	Not flagged
Evaluate a regex-filtered flag Shard: flags-single	Allocated memory per operation	18.91 KiB/op	19.24 KiB/op	1.7% more memory	Not flagged
Evaluate a regex-filtered flag Shard: flags-single	Allocations per operation	203.77 count/op	210.77 count/op	3.4% more	Not flagged
Evaluate a regex-filtered flag Shard: flags-single	Time per operation	11.45 µs/op	11.45 µs/op	<0.1% more time	Not flagged
Evaluate a regex-filtered flag Shard: flags-single	Process CPU time per operation	11.45 µs/op	11.45 µs/op	<0.1% more time	Not flagged
Evaluate a rollout flag Shard: flags-single	Allocated memory per operation	13.72 KiB/op	14.05 KiB/op	2.4% more memory	Not flagged
Evaluate a rollout flag Shard: flags-single	Allocations per operation	153.76 count/op	160.76 count/op	4.6% more	Not flagged
Evaluate a rollout flag Shard: flags-single	Time per operation	9.16 µs/op	9.11 µs/op	0.5% less time	Not flagged
Evaluate a rollout flag Shard: flags-single	Process CPU time per operation	9.16 µs/op	9.11 µs/op	0.5% less time	Not flagged
Evaluate a multivariate flag Shard: flags-single	Allocated memory per operation	14.89 KiB/op	15.22 KiB/op	2.2% more memory	Not flagged
Evaluate a multivariate flag Shard: flags-single	Allocations per operation	170.77 count/op	177.77 count/op	4.1% more	Not flagged
Evaluate a multivariate flag Shard: flags-single	Time per operation	10.08 µs/op	10.04 µs/op	0.4% less time	Not flagged
Evaluate a multivariate flag Shard: flags-single	Process CPU time per operation	10.08 µs/op	10.04 µs/op	0.4% less time	Not flagged

Values are medians of paired measurements. Not flagged does not mean unchanged. Unavailable means a measurement could not be provided, not zero. Estimated allocation measurements retain their instrumentation uncertainty. RSS measures the whole process, not just SDK allocations; observed peaks are sampled checkpoints. Paired ranges in the diagnostics describe variability, not confidence intervals.

Each shard uses its own runner; compare Before PR against With PR within a row, not across shards.

Compared 4f60c3392bc93a53ede2e61d9079447ce0d950a8 → 80563b2fb1dc6dc874d36e30eb673202f7269d1e.

Full report and diagnostics

102 metric rows; 102 rows shown in all results. Omitted or shortened data remain in comparison.json, including per-runner environments, resource limits, measurement methods, samples and checkpoints.

[veria-ai]

PR overview

No security issues were found in this pull request.

Reviews

• 80563b2 — No new issues