[dependabot]

Bumps PostHog/posthog-sdk-test-harness/.github/workflows/test-sdk-action.yml from 1.15.0 to 1.19.0.

Release notes

Sourced from PostHog/posthog-sdk-test-harness/.github/workflows/test-sdk-action.yml's releases.

1.19.0

Minor changes

• 65b0eb7 Include migrated YAML-parity cases in the SDK acceptance suite while preserving their capability-based selection. — Thanks @​dustinbyrne!

1.18.0

Minor changes

• a6e61f6 Generate the per-field feature flag rules v2 response negatives from the presence matrix instead of committing them. For every row of rules/response_presence.json, the meta-tests build a missing-field case for has_experiment and each required field, a forbidden-field case for each forbidden field, and a wrong condition index case from the row's template, and check each against the same validation layer, keyword and instance path the committed cases declared. A mutation that another matrix row accepts is not a negative. 76 of the 77 committed cases that the generator reproduces exactly (responses.<row>_missing_<field>, responses.<row>_forbidden_<field> and responses.<row>_wrong_condition_index) are removed from fixtures/wire/responses.json; responses.holdout_missing_holdout_id stays so that the committed set keeps a presence-layer required failure. The hand-made invalid cases, all valid cases and all templates stay. responses.json now has 25 valid and 49 invalid cases, so consumers that pin its case counts must update them when they re-vendor. The removal ships in contract package 5.0.0. — Thanks @​andehen!

1.17.0

Minor changes

• a599572 Remove the feature flag rules v2 wire fixture sets that no consumer runs: fixtures/wire/calls.json, call_transport.json, exposures.json, exposure_transport.json, definitions_feed.json, errors.json and warnings.json, together with their manifest.json entries and the meta-tests that only checked them. The definitions.json, responses.json and readers.json fixture sets are unchanged. All schemas stay byte-identical and published as specifications, including the $feature_flag_called context, experiment exposure, management warning, management error and definitions feed schemas; the meta-tests still check each one as a valid JSON Schema 2020-12 document. Fixture sets are added together with the first consumer that runs them. The removal ships in contract package 4.0.0. — Thanks @​andehen!

1.16.0

Minor changes

• 429c266 Use one version for the feature flag rules v2 contract (contract package 3.0.0). manifest.json now holds only contract (name, version, config_version) and artifacts entries with path and kind, plus fixture_id, expected and expected_failure on config fixtures. SHA256SUMS and its checksum script are removed, as are the per-component versions: case_ids, per-artifact version, schema and component in the manifest, fixture_version in the wire fixture files, registry_version in the literal registry and version in the presence matrix. Consumers re-vendoring the contract must stop reading these, take case IDs from the corpus and fixture files, and pin the harness release or commit instead of the checksum index. Schema bytes, schema $ids and all fixture and corpus expectations are unchanged. — Thanks @​andehen!

Changelog

Sourced from PostHog/posthog-sdk-test-harness/.github/workflows/test-sdk-action.yml's changelog.

posthog-sdk-test-harness

1.22.0 — 2026-10-26

Minor changes

• 0bc2c7c Slim the feature flag rules v2 experiment rule evaluation corpus (contract package 7.0.0). The corpus drops about 38% of its bytes and loses no expectation: every removed field follows from the fields that remain.

  • corpus/v2_experiment_evaluation.json gains defaults with the flag and context every case shares. A case no longer carries flag, and its context holds only the fields that differ from the default context.
  • Cases no longer carry legacy; consumers derive it from expected.value with the typed value corpus table.
  • hash_evidence entries keep use, prefix, identifier, salt and hash01_binary64_hex. The SHA-1 intermediates move out to corpus/hash_sha1_60_v1.json, which already pins them, and the threshold or boundaries follow from the rule.
  • Removed cases: v2_experiment.assignment.fifty_fifty_test, an exact copy of v2_experiment.ordering.split, and seven parser rows that repeat a config fixture: parser.experiment_id, parser.holdout_id, parser.missing_paused, parser.one_variant, parser.twenty_one_variants, parser.variant_key_characters and parser.variant_value_type. The corpus has 73 cases. — Thanks @​andehen!

1.21.0 — 2026-10-26

Minor changes

• 2c3a2ae Rename the feature flag rules v2 /flags response record from v3 to v4 (contract package 7.0.0). Clients already send v=3 and arbitrary v values to /flags, so the service selects the typed record only for an exact v=4, and the contract now names the record after that request version. The record's shape and every expectation are unchanged.

  • Schemas: schemas/flags_response_v3.schema.json is renamed to schemas/flags_response_v4.schema.json, with $id https://posthog.com/contracts/feature_flag_rules_v2/flags_response_v4.schema.json and title POST /flags?v=4 response (producer contract). schemas/flags_response_v3_presence.schema.json is renamed to schemas/flags_response_v4_presence.schema.json, with the matching $id, title and $ref. These $id changes are a deliberate exception to the stable-$id rule, made in this major version.
  • Fixtures: fixtures/wire/responses.json and fixtures/wire/readers.json point at the renamed producer schema. The reader cases readers.v3_value_takes_precedence and readers.v1_at_v3 are renamed to readers.v4_value_takes_precedence and readers.v1_at_v4, with the same expectations.
  • Manifest and README follow the renames. — Thanks @​andehen!

1.20.0 — 2026-10-21

Minor changes

• 34fd648 Add the feature flag rules v2 experiment rule evaluation corpus and drop the experiment and holdout references from v2 experiment rules (contract package 6.0.0). An experiment links to its rule from the Experiment row, so the rule carries no experiment reference, and every variant assignment by an experiment rule reports experiment_split, whether or not an experiment links to the rule. A holdout belongs to its rule: it holds only a seed and an exclusion percentage, and a holdout result is identified by its rule ID and the holdout reason.

  • Config schema: experiment_id is no longer a rule field and id is no longer a holdout field, so a rule or holdout that carries one is rejected. config.invalid.null_experiment_id is removed; config.valid.experiment_rule_local_holdout, config.invalid.experiment_rule_with_experiment_id and config.invalid.holdout_with_id are added; the other experiment rule fixtures drop experiment_id and holdout.id and keep their expectations.
  • Corpus: corpus/v2_experiment_evaluation.json (81 cases in the ordering, assignment, values, white_box and parser families) and its companion schema.
  • Wire: no config version 2 record carries experiment_id, and no record or event carries a holdout ID. The producer schema forbids experiment_id on v2 records, no longer requires it on a split, and has no holdout_id; the presence rows for experiment rules forbid experiment_id; the event schemas have no $feature_flag_holdout_id, and $experiment_exposure has no $feature_flag_experiment_id. Version 1 records keep experiment_id. The wire templates drop both fields from v2 records and id from the definitions holdout. Removed cases: responses.split_wrong_type_experiment_id, responses.experiment_split_contradicted_has_experiment, responses.rollout_miss_percentage_rollout_wrong_rule_type, responses.rollout_miss_experiment_wrong_rule_type, responses.holdout_missing_holdout_id, responses.v1_forbidden_holdout_id and readers.partial_split_missing_experiment_id. Added cases: responses.experiment_split_flag_without_experiment, responses.holdout_missing_rule_id (the committed presence-layer required failure) and readers.split_extra_experiment_id. responses.json now has 26 valid and 44 invalid cases.
  • Registry: experiment_id leaves fields, and experiment_id and holdout_id leave openfeature_mapping.flag_metadata_keys. — Thanks @​andehen!

1.19.0 — 2026-10-20

Minor changes

• 65b0eb7 Include migrated YAML-parity cases in the SDK acceptance suite while preserving their capability-based selection. — Thanks @​dustinbyrne!

1.18.0 — 2026-10-20

Minor changes

• a6e61f6 Generate the per-field feature flag rules v2 response negatives from the presence matrix instead of committing them. For every row of rules/response_presence.json, the meta-tests build a missing-field case for has_experiment and each required field, a forbidden-field case for each forbidden field, and a wrong condition index case from the row's template, and check each against the same validation layer, keyword and instance path the committed cases declared. A mutation that another matrix row accepts is not a negative. 76 of the 77 committed cases that the generator reproduces exactly (responses.<row>_missing_<field>, responses.<row>_forbidden_<field> and responses.<row>_wrong_condition_index) are removed from fixtures/wire/responses.json; responses.holdout_missing_holdout_id stays so that the committed set keeps a presence-layer required failure. The hand-made invalid cases, all valid cases and all templates stay. responses.json now has 25 valid and 49 invalid cases, so consumers that pin its case counts must update them when they re-vendor. The removal ships in contract package 5.0.0. — Thanks @​andehen!

1.17.0 — 2026-10-19

Minor changes

... (truncated)

Commits

• 129fbb4 chore: Release v1.19.0 [skip ci]
• 65b0eb7 feat(v2): unify migrated and canonical SDK acceptance (#88)
• 3957869 chore: Release v1.18.0 [skip ci]
• a6e61f6 test(flags): generate rules v2 presence negatives from the matrix (#86)
• 47178f6 chore(deps): bump the github-actions group across 1 directory with 4 updates ...
• 8ca75c4 chore: Release v1.17.0 [skip ci]
• a599572 refactor(flags): drop rules v2 wire fixture sets that nothing runs (#85)
• 27ae744 chore: Release v1.16.0 [skip ci]
• 429c266 refactor(flags): use one version for the rules v2 contract (#84)
• 4af3205 ci: run the harness unit tests (#83)
• See full diff in compare view

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

---

Dependabot commands and options

You can trigger Dependabot actions by commenting on this PR:

• @dependabot rebase will rebase this PR
• @dependabot recreate will recreate this PR, overwriting any edits that have been made to it
• @dependabot show <dependency name> ignore conditions will show all of the ignore conditions of the specified dependency
• @dependabot ignore this major version will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)
• @dependabot ignore this minor version will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)
• @dependabot ignore this dependency will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)

[github-actions]

posthog-android Compliance Report

Date: 2026-10-27 06:22:00 UTC
Duration: 118423ms

✅ All Tests Passed!

46/46 tests passed

---

Capture Tests

✅ 29/29 tests passed

View Details

Test	Status	Duration
Format Validation.Event Has Required Fields	✅	349ms
Format Validation.Event Has Uuid	✅	34ms
Format Validation.Event Has Lib Properties	✅	37ms
Format Validation.Distinct Id Is String	✅	29ms
Format Validation.Token Is Present	✅	29ms
Format Validation.Custom Properties Preserved	✅	30ms
Format Validation.Event Has Timestamp	✅	25ms
Retry Behavior.Retries On 503	✅	7031ms
Retry Behavior.Does Not Retry On 400	✅	4025ms
Retry Behavior.Does Not Retry On 401	✅	4027ms
Retry Behavior.Respects Retry After Header	✅	7031ms
Retry Behavior.Implements Backoff	✅	17036ms
Retry Behavior.Retries On 500	✅	7021ms
Retry Behavior.Retries On 502	✅	7019ms
Retry Behavior.Retries On 504	✅	7020ms
Retry Behavior.Max Retries Respected	✅	17036ms
Deduplication.Generates Unique Uuids	✅	43ms
Deduplication.Preserves Uuid On Retry	✅	7012ms
Deduplication.Preserves Uuid And Timestamp On Retry	✅	12032ms
Deduplication.Preserves Uuid And Timestamp On Batch Retry	✅	7019ms
Deduplication.No Duplicate Events In Batch	✅	42ms
Deduplication.Different Events Have Different Uuids	✅	26ms
Compression.Sends Gzip When Enabled	✅	18ms
Batch Format.Uses Proper Batch Structure	✅	20ms
Batch Format.Flush With No Events Sends Nothing	✅	12ms
Batch Format.Multiple Events Batched Together	✅	32ms
Error Handling.Does Not Retry On 403	✅	4019ms
Error Handling.Does Not Retry On 413	✅	4022ms
Error Handling.Retries On 408	✅	5027ms

Feature_Flags Tests

✅ 17/17 tests passed

View Details

Test	Status	Duration
Request Payload.Request With Person Properties Device Id	✅	40ms
Request Payload.Flags Request Uses V2 Query Param	✅	30ms
Request Payload.Flags Request Hits Flags Path Not Decide	✅	26ms
Request Payload.Flags Request Omits Authorization Header	✅	25ms
Request Payload.Token In Flags Body Matches Init	✅	22ms
Request Payload.Groups Round Trip	✅	21ms
Request Payload.Groups Default To Empty Object	✅	23ms
Request Payload.Disable Geoip False Propagates As Geoip Disable False	✅	21ms
Request Payload.Disable Geoip Omitted Defaults To False	✅	29ms
Request Payload.Flag Keys To Evaluate Contains Only Requested Key	✅	31ms
Request Lifecycle.No Flags Request On Init Alone	✅	12ms
Request Lifecycle.No Flags Request On Normal Capture	✅	18ms
Request Lifecycle.Two Flag Calls Produce Two Remote Requests	✅	41ms
Request Lifecycle.Mock Response Value Is Returned To Caller	✅	24ms
Retry Behavior.Retries Flags On 502	✅	328ms
Retry Behavior.Retries Flags On 504	✅	325ms
Side Effect Events.Get Feature Flag Captures Feature Flag Called Event	✅	26ms

[dustinbyrne]

Reviewed the reusable compliance-workflow pin update at this commit. Verified upstream 1.19.0 SHA/tag mapping and nested upload-artifact v7.0.2 mapping. Existing harness image 0.10.0 and workflow inputs/permissions remain unchanged; no SDK runtime or dependency changes. Required Build Job and organization Semgrep pass; compliance capture/flags check passes. No edits or lockfile changes.