feat(programs): B2 surface shell — runProgram types and stub signatures by gewenyu99 · Pull Request #1307 · PostHog/wizard

gewenyu99 · 2026-09-22T21:20:59Z

Scope: New surface only: types and stubs, no plumbing. Review the contract.

Related: #1320

This adds the runProgram surface as a shell. It declares what a caller passes in and gets back, and runProgram throws until #1308 fills it in. Nothing calls it yet.

Stack: #1306 moves → #1307 surface shell → #1308 plumbing → #1357 runner context → #1309 docs → #1363 stream retry.

What it declares

run-program.ts: ProgramInput, ProgramSettings, ProgramOptions, ProgramRunOutcome and a stub runProgram. Each field has a one-line comment.
program-store.ts: ProgramProgress, ProgramInvocationData and the ProgramStore shape. It imports through the @agent/types and @shared/api aliases.
credentials.ts: ResolvedProgramCredentials and CredentialsProvider.
DiscoveredFeature: it now lives in src/shared/discovered-feature.ts, so programs can name it without the session. wizard-session.ts re-exports it, so its readers keep their import path.
Agent types: the run options agentic detection needs, prompt, collectTranscript, requestRemark and scanReport, plus the activity progress event.

Checks at 68a321cd: pnpm typecheck, pnpm lint (0 errors), pnpm vitest run (3,405 tests, the same as #1306, since the shell adds none) and pnpm test:arch (11 tests) pass.

Created with PostHog Desktop

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

Store authentication, detection, and composition for a single program invocation. Record actual finished agent results separately from projected progress while preserving existing result order. Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

Move scan consent and cleanup registration into UI-free shared leaves. Decouple package-manager file operations from CLI setup and assert the full runProgram import closure. Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

github-actions · 2026-09-22T21:21:12Z

🧙 Wizard CI

Run the Wizard CI and test your changes against wizard-workbench example apps by replying with a GitHub comment using one of the following commands:

Test all apps:

/wizard-ci all

Test all apps in a directory:

/wizard-ci ai-observability
/wizard-ci basic-integration
/wizard-ci mcp-analytics
/wizard-ci replay-vision
/wizard-ci revenue
/wizard-ci self-driving
/wizard-ci warehouse
/wizard-ci warehouse-seeded

Test an individual app:

/wizard-ci ai-observability/anthropic
/wizard-ci ai-observability/google-adk
/wizard-ci ai-observability/groq

Show more apps

/wizard-ci ai-observability/manual-capture
/wizard-ci ai-observability/openai
/wizard-ci ai-observability/openai-agents
/wizard-ci ai-observability/opentelemetry
/wizard-ci ai-observability/vercel-ai
/wizard-ci basic-integration/android
/wizard-ci basic-integration/angular
/wizard-ci basic-integration/astro
/wizard-ci basic-integration/django
/wizard-ci basic-integration/fastapi
/wizard-ci basic-integration/flask
/wizard-ci basic-integration/flutter
/wizard-ci basic-integration/javascript-node
/wizard-ci basic-integration/javascript-web
/wizard-ci basic-integration/laravel
/wizard-ci basic-integration/next-js
/wizard-ci basic-integration/nuxt
/wizard-ci basic-integration/python
/wizard-ci basic-integration/rails
/wizard-ci basic-integration/react-native
/wizard-ci basic-integration/react-router
/wizard-ci basic-integration/sveltekit
/wizard-ci basic-integration/swift
/wizard-ci basic-integration/tanstack-router
/wizard-ci basic-integration/tanstack-start
/wizard-ci basic-integration/vue
/wizard-ci mcp-analytics/custom-dispatcher
/wizard-ci mcp-analytics/typescript-sdk
/wizard-ci replay-vision/javascript-node
/wizard-ci replay-vision/next-js
/wizard-ci replay-vision/react-vite
/wizard-ci revenue/stripe
/wizard-ci self-driving/astro
/wizard-ci self-driving/fastapi
/wizard-ci self-driving/nuxt
/wizard-ci self-driving/react-router
/wizard-ci self-driving/sveltekit
/wizard-ci warehouse/monorepo-env
/wizard-ci warehouse/multi-source-next
/wizard-ci warehouse/stripe-node
/wizard-ci warehouse/zero-source
/wizard-ci warehouse-seeded/next-stripe
/wizard-ci warehouse-seeded/next-stripe-declined

Test against a Context Mill branch:

/wizard-ci all context-mill:my-branch

Add context-mill:<branch> to any command above to pin the Context Mill branch. It defaults to main.

Results will be posted here when complete.

gewenyu99 · 2026-09-22T21:21:24Z

B2 work in progress. Local verification on this head: pnpm typecheck and 42 focused architecture, callable-host, and legacy-adapter tests pass. Existing signed commits preserve the repository trailers. The watcher migration and full snapshot parity review are still in progress; follow-up commits will update this draft.

gewenyu99 · 2026-09-22T21:21:43Z

/wizard-ci-snapshots basic-integration/javascript-node/express-todo

wizard-ci-bot · 2026-09-22T21:22:16Z

🧙 Wizard CI Results

Trigger ID: 4dc8421
Workflow: View run

App	Confidence	PR	YARA
`basic-integration/javascript-node/express-todo`	N/A	Failed (logs)

Configuration

Setting	Value
Wizard ref	`posthog/functional-b-integration`
Context Mill ref	`main`
PostHog ref	`master`

Search for trigger ID 4dc8421 in wizard-workbench PRs.

Own artifact watchers and CI inference auth in programs, enforce composition gates, and clean run-installed skills on every failed path. Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

gewenyu99 · 2026-09-22T21:37:41Z

B2 follow-up at e334fe98: callable audit and event-plan watchers now write invocation data and release on every exit; self-driving requires an affirmative GitHub gate; failed parent runs remove Wizard-installed child skills; CI inference auth reaches the host without agent-global setup. Validation: pnpm typecheck, pnpm test:programs (768/768), pnpm test:arch (13/13), pnpm lint (0 errors), pnpm build and smoke passed. The two localhost skill-download tests time out inside the sandbox but pass together outside it (109/109). Snapshot parity remains B3 and is being reviewed separately.

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

gewenyu99

🥸 Reviewed by NotVincent, totally not Vincent. The real Vincent will review it separately. Probably slop, please disregard.

gewenyu99 · 2026-09-22T23:13:08Z

+
+function fingerprintSkill(skillDir: string): string {
+  const digest = createHash('sha256');
+  const files = fg.sync(SKILL_TEXT_GLOB, {


Here's the potential issue: The scan exists to stop a poisoned project skill reaching the agent, but it only matches the exact name SKILL.md, while macOS and Windows filesystems resolve names case-insensitively.

repo ships .claude/skills/x/SKILL.MD with injection text -> scan globs for SKILL.md and reads nothing -> reports clean -> SDK opens SKILL.md, gets the uppercase file -> skill loads unscanned

Suggested fix: Match case-insensitively on both skill globs, here and in the YARA file scan, and add a SKILL.MD fixture that expects the poisoned verdict.

NotVincent here. This no longer applies to B: the project skill scan at load was cut, and the YARA file scan's case fix is its own PR off main, #1344.

gewenyu99 · 2026-09-22T23:13:08Z

+  workingDirectory: string,
+  triageProvider: LLMProvider | undefined,
+): Promise<ProjectSkillFinding[]> {
+  const root = path.join(workingDirectory, '.claude', 'skills');


Here's the potential issue: The scan covers only the install directory's .claude/skills, but the SDK loads .claude/skills from every directory up to the git root.

monorepo root ships a poisoned skill, user runs --install-dir apps/web -> scan finds apps/web/.claude/skills empty -> reports clean -> SDK walks up and loads the root skill -> skill loads unscanned

Suggested fix: Scan every .claude/skills from the working directory up to the git root, stopping at the home directory.

NotVincent here. This no longer applies: B cut the project skill scan at load, so there is no scan root left to widen.

gewenyu99 · 2026-09-22T23:13:08Z

        emitStepEvents: config.trackStepProgress ?? false,
        resolveStepKey: config.resolveStepKey,
        triageProvider: boot.triageProvider,
+        signal: inputs.signal,


Here's the potential issue: Cancellation is meant to stop the model, but nothing checks that either real harness forwards the host signal into the SDK query or the Pi session.

someone drops the signal from both harnesses -> all 3285 tests stay green -> host cancels -> SDK query keeps running and spending inference -> runner still reports Aborted

Suggested fix: Call both harnesses with a live AbortController, and assert the signal reaches the SDK query and that the Pi session's abort() fires.

NotVincent here. B no longer changes how the harnesses forward the signal, so these tests moved to #1345 off main, which checks that the Pi session aborts on a live host cancel.

gewenyu99 · 2026-09-22T23:13:08Z

+  const findings: ProjectSkillFinding[] = [];
+  for (const entry of fs.readdirSync(root, { withFileTypes: true })) {
+    const skillDir = path.join(root, entry.name);
+    if (!entry.isDirectory() && !fs.statSync(skillDir).isDirectory()) continue;


Here's the potential issue: A broken symlink the SDK would skip stops the whole run as a security failure. Separately, no test uses a symlinked skill, and this statSync branch is the only thing that lets one into the scan.

dotfiles-linked skill whose target moved -> statSync throws ENOENT -> caller maps it to YARA_VIOLATION -> every run fails with "Security check could not scan project skills"

Suggested fix: Stat with throwIfNoEntry: false and skip entries that aren't directories. Add a test that symlinks a poisoned skill directory and expects the poisoned verdict, so this branch can't be narrowed to real directories unnoticed.

NotVincent here. This no longer applies: B cut the project skill preflight, so a broken skill link can no longer fail the run.

gewenyu99 · 2026-09-22T23:13:08Z

+      if (!descriptor) continue;
+      if ('value' in descriptor) {
+        const value: unknown = descriptor.value;
+        descriptor.value = structuredClone(value);


Here's the potential issue: The outcome should carry the run's real failure, but cloning every property of the error throws for common error types.

run crashes with an AxiosError -> finish deep-clones config.transformRequest, a function -> DataCloneError -> runProgram rejects with the clone error -> user and analytics see DataCloneError, not the crash

Suggested fix: Keep the error instance by reference, or fall back to the original value when cloning fails.

NotVincent here. The store no longer clones the run result: finish keeps it by reference, and the store test checks that the settled result is the same object, in #1308:

wizard/src/programs/program-store.ts

Lines 124 to 126 in f2e42ac

finish: (result) => {

run.result = result;

},

gewenyu99 · 2026-09-22T23:13:09Z

+      continue;
+    }
+
+    const reason = await scanInstalledSkill(


Here's the potential issue: The skill the Wizard just installed and scanned gets scanned again here.

linear run installs its skill -> download scan passes but records nothing -> preflight misses its cache -> second YARA pass, and a second triage call if a chunk is flagged

Suggested fix: Record the clean verdict after the install scan.

NotVincent here. This no longer applies: B cut the project skill preflight, so nothing scans the installed skill a second time.

gewenyu99 · 2026-09-22T23:13:09Z

    sequence: (process.env.SNAP_SEQUENCE || undefined) as Sequence | undefined,
    model: process.env.SNAP_MODEL || undefined,
  });
+  store.setInferenceAuth(


Here's the potential issue: tui-host used to report a missing CI token as a readable run failure. It now reads the token file at startup and dies first.

start tui-host without WIZARD_CI_GATEWAY_TOKEN_FILE -> FATAL before the control socket serves -> MCP client sees a dead socket -> detection-only route and screen-only commands stop working

Suggested fix: Read the token file on first use rather than at startup.

NotVincent here. This no longer applies: B cut the CI bearer wiring, and tui-host no longer reads a CI token file.

gewenyu99 · 2026-09-22T23:13:09Z

      programId,
+      inferenceAuth:
+        options.inferenceAuth ??
+        session.inferenceAuth ??


Here's the potential issue: This fallback is what gives --ci detection the CI bearer, and nothing tests it.

someone deletes session.inferenceAuth here -> suite stays green -> CI detection mints with the phx key again

Suggested fix: Assert detection passes the session's provider to the agent.

NotVincent here. This no longer applies: B cut the CI bearer wiring, so detection has no session.inferenceAuth fallback left.

gewenyu99 · 2026-09-22T23:13:09Z

        // out of the TUI's startup graph.
        programId: store.analyticsProgramId,
+        inferenceAuth:
+          store.session.inferenceAuth ??


Here's the potential issue: The store half of the CI bearer has no test.

delete this fallback or the store setter in runNonInteractive -> suite stays green -> tui-host mcp-tutorial mints again

Suggested fix: Assert the store's provider after a CI start and its use here.

NotVincent here. This no longer applies: B cut the CI bearer wiring, so the TUI store no longer carries a CI provider.

gewenyu99 · 2026-09-22T23:13:09Z

  rerankIds?: readonly string[];
  /** Streaming activity callback for the UI. */
  onEvent?: DetectEvent;
+  inferenceAuth?: import('@agent/types').InferenceAuthProvider;


Here's the potential issue: This option adds a third source for detection auth, and nothing sets it.

every caller omits it -> the branch is dead -> readers reason about three precedence levels where two are real

Suggested fix: Delete it until a caller needs it.

NotVincent here. The option is gone: AgenticDetectOptions has no auth field, and detection passes the session's credentials to runAgent:

wizard/src/programs/detection/agentic.ts

Lines 129 to 153 in f2e42ac

export type AgenticDetectOptions = {

/**

* Categories to classify each project into. Order matters: list targets by

* priority, most specific first. When a project could match more than one,

* the agent keeps the earliest — so the caller encodes precedence purely

* through ordering (e.g. a bundler target before a generic framework target).

*/

targets: readonly DetectTarget[];

/** The program this scan bills to. Required, not optional: the scan drives a

* real agent through the gateway, and a caller that forgets leaves that

* spend unattributed. */

programId: string;

/** One short clause describing what the scan is for (frames the prompt). */

purpose?: string;

/** Ask the agent to label exactly one project `recommended` (the main client app). Off by default. */

recommend?: boolean;

/**

* Target ids whose technologies co-occur in one manifest (e.g. the JS

* family). Among these, the enumeration's priority winner overrides the

* model's pick; elsewhere a valid pick always wins — see coerceAgenticReport.

*/

rerankIds?: readonly string[];

/** Streaming activity callback for the UI. */

onEvent?: DetectEvent;

};

@programs

Bring in the adapter move to src/lib/runners/run-program-agent.ts. B2's adapter changes land on the new path, its @programs imports replace the relative ones, and known-violations.json lists the moved adapter's edges. Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

Set the tree to B1's, then declare the new surfaces with their final names, paths and signatures, taken from B3 at 50fe9e4 with the bodies removed. - src/programs/run-program.ts: ProgramOverrides, WizardFlagSnapshot, ProgramSettings, ProgramInput, ProgramOptions, ProgramRunOutcome, and a runProgram that throws "runProgram: not implemented". - src/programs/program-store.ts: the progress, diagnostic, invocation-data and settled-run types, and ProgramStore's public methods, each throwing. - src/programs/credentials.ts: ResolvedProgramCredentials, CredentialsProvider, and createPosthogInferenceAuthProvider, throwing. - src/programs/host-capabilities.ts: ProgramCiHost and ProgramRunHost, with AuthProjection in authenticate.ts and ProgramCompletionContext in program-run.ts. - The programs entry exports runProgram and createPosthogInferenceAuthProvider as lazy loaders, and the type entry exports the new types. - The agent declares InferenceAuthProvider, the collectTranscript, requestRemark and prompt run-definition fields, RunConfig.scanReport, RunInput.inferenceAuth (optional until B3 supplies it), RunSnapshot.transcriptTail, and the activity progress event, which the UI reducer ignores. Nothing calls the shell: the adapter, the CLI and every doc stay as B1 has them, and no tests are added. Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

Take B1's tree as it stands, the original B1 moves only. The surface shell is declared again on top of it. Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

@programs

…nal B1 Declare the new surfaces with their final names and signatures, taken from B3 at f1b30d7 with the bodies removed, against B1's layout. - src/programs/run-program.ts: ProgramOverrides, WizardFlagSnapshot, ProgramSettings, ProgramInput, ProgramOptions, ProgramRunOutcome, and a runProgram that throws "runProgram: not implemented". - src/programs/program-store.ts: the progress, diagnostic, invocation-data and settled-run types, and ProgramStore's public methods, each throwing. - src/programs/credentials.ts: ResolvedProgramCredentials and CredentialsProvider. - @programs exports runProgram, and @programs/types exports the new types. - The agent declares the prompt, collectTranscript and requestRemark run-definition fields, RunConfig.scanReport, RunSnapshot.transcriptTail, and the activity progress event, which the UI reducer ignores. @agent/types exports ResolvedBinding and RunHooks for the shell. B1 keeps DiscoveredFeature in the legacy session, so known-violations.json records run-program.ts's type import of it until B3 moves it to shared. Inference auth, the deferred skill commit, the watcher fields, the lazy @agent entry and the host capabilities stay in B3. Nothing calls the shell: the adapter, the CLI and every doc stay as B1 has them, and no tests are added. Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

Records #1307's shell as this branch's base. The tree is unchanged here; the next commits return it to the shell's tree and add the plumbing. Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

Set the tree to #1307's shell at a467ba9, so the commits after this add only the plumbing the shell needs. Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

gewenyu99 · 2026-09-25T15:53:48Z

this isn't hooked up yet in this PR, we will move this in a later PR

Implemented in #1308: program-store.ts (at f550e227).

johncwaters

Note

Automated review. Not written by a human.

Verdict: APPROVE WITH NITS

See inline comments. The real John is already reading with physical eyes.

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

…-store imports DiscoveredFeature moves to src/shared/discovered-feature.ts and the session re-exports it, so run-program.ts no longer reaches the session and its allowlist row goes. program-store.ts imports through @agent/types and @shared/api. Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

gewenyu99 · 2026-09-25T17:53:30Z

NotVincent here. This express-todo failure was a harness bug, not this PR: the run completed and reached keep-skills in 26 frames, but result.json stops at the outro, which #1340 fixes off main. The trigger 4dc8421 also ran on an old head: https://github.com/PostHog/wizard-workbench/actions/runs/35786186197/job/106943771140

gewenyu99 added 18 commits September 22, 2026 17:15

refactor(skills): separate download files from agent scan

963403d

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

fix(agent): scan project skills before SDK load

9fc46f1

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

refactor(programs): project agent progress into invocation state

36f3805

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

refactor(programs): add explicit inference credentials providers

111ffba

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

refactor(programs): own binding policy and experiment selection

077cec5

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

fix: clean run-installed skills on failed Wizard exits

accaeec

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

refactor(programs): resolve dynamic runs from explicit data

a24cc40

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

feat: run no-agent programs headlessly

fb8b36f

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

refactor(programs): resolve dynamic integration recipes from data

128c5cf

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

refactor(agent): use caller inference auth in both harnesses

2fbbdcc

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

feat(programs): add callable runProgram host

36295c9

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

refactor(programs): load callable program config without TUI

b9b978f

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

feat(programs): compose runs and expose invocation outcomes

e9360b6

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

refactor(programs): route legacy agent runs through callable host

28411b6

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

feat(agent): propagate host cancellation through active runs

80785b1

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

feat(programs): honor host cancellation before and during runs

33e55dd

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

refactor(programs): finish callable host lifecycle and cleanup

e334fe9

Own artifact watchers and CI inference auth in programs, enforce composition gates, and clean run-installed skills on every failed path. Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

refactor(programs): own gateway authentication outside agent

49f75e6

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

gewenyu99 commented Sep 22, 2026

View reviewed changes

gewenyu99 added 4 commits September 22, 2026 21:27

Merge A3 and B1 review fixes into B2

4f519e8

fix callable program host boundaries

0947e09

Merge B2 callable host review fixes

72d3aff

fix: secure project skill preflight and run cleanup

4fcdb1f

gewenyu99 changed the base branch from posthog/functional-b2-move to posthog/functional-b1-pr September 24, 2026 20:58

gewenyu99 added 2 commits September 24, 2026 17:07

gewenyu99 changed the title ~~feat(programs): B2b new code — callable runProgram host~~ feat(programs): B2 surface shell — runProgram types and stub signatures Sep 24, 2026

gewenyu99 mentioned this pull request Sep 24, 2026

feat(programs): B3 plumbing and tests — runProgram, adapter wiring, detection via runAgent #1308

Merged

gewenyu99 added 2 commits September 24, 2026 18:28

Merge posthog/functional-b1-pr into posthog/functional-b-integration

42a379d

Take B1's tree as it stands, the original B1 moves only. The surface shell is declared again on top of it. Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

gewenyu99 added 6 commits September 24, 2026 19:50

refactor(programs): call it runProgram, keep host for callers

1725cc3

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

docs(programs): describe each runProgram field inline

b3a38c6

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

refactor(programs): call it the runner, not the host

1e84cbc

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

Merge posthog/functional-b1-pr into posthog/functional-b-integration

9e6a0e7

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

Merge posthog/functional-b1-pr into posthog/functional-b-integration

67cb4da

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

docs(programs): correct four runProgram field comments

53574fe

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

gewenyu99 added this pull request to stack #1362 September 25, 2026 15:16

gewenyu99 commented Sep 25, 2026

View reviewed changes

gewenyu99 marked this pull request as ready for review September 25, 2026 15:54

gewenyu99 requested a review from a team as a code owner September 25, 2026 15:54

johncwaters reviewed Sep 25, 2026

View reviewed changes

gewenyu99 added 3 commits September 25, 2026 12:39

Merge posthog/functional-b1-pr into posthog/functional-b-integration

a87c695

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

Merge posthog/functional-b1-pr into posthog/functional-b-integration

68a321c

Generated-By: PostHog Desktop Task-Id: d14e92bb-6ee1-49b5-8502-39cb80079589

johncwaters approved these changes Sep 25, 2026

View reviewed changes

This was referenced Sep 25, 2026

refactor(programs): B1 move-only — program code into src/programs #1306

Merged

docs(programs): B5 docs-only — runProgram and developer interfaces #1309

Merged

gewenyu99 merged commit 92e6fe5 into main Sep 25, 2026
19 checks passed

gewenyu99 deleted the posthog/functional-b-integration branch September 25, 2026 23:58

releaser-wizard Bot mentioned this pull request Sep 25, 2026

chore(main): release 2.79.0 #1365

Merged

	export type AgenticDetectOptions = {
	/**
	* Categories to classify each project into. Order matters: list targets by
	* priority, most specific first. When a project could match more than one,
	* the agent keeps the earliest — so the caller encodes precedence purely
	* through ordering (e.g. a bundler target before a generic framework target).
	*/
	targets: readonly DetectTarget[];
	/** The program this scan bills to. Required, not optional: the scan drives a
	* real agent through the gateway, and a caller that forgets leaves that
	* spend unattributed. */
	programId: string;
	/** One short clause describing what the scan is for (frames the prompt). */
	purpose?: string;
	/** Ask the agent to label exactly one project `recommended` (the main client app). Off by default. */
	recommend?: boolean;
	/**
	* Target ids whose technologies co-occur in one manifest (e.g. the JS
	* family). Among these, the enumeration's priority winner overrides the
	* model's pick; elsewhere a valid pick always wins — see coerceAgenticReport.
	*/
	rerankIds?: readonly string[];
	/** Streaming activity callback for the UI. */
	onEvent?: DetectEvent;
	};

Conversation

gewenyu99 commented Sep 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions Bot commented Sep 22, 2026

🧙 Wizard CI

Uh oh!

gewenyu99 commented Sep 22, 2026

Uh oh!

gewenyu99 commented Sep 22, 2026

Uh oh!

wizard-ci-bot Bot commented Sep 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🧙 Wizard CI Results

Configuration

Uh oh!

gewenyu99 commented Sep 22, 2026

Uh oh!

gewenyu99 left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

johncwaters left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

gewenyu99 commented Sep 25, 2026

Uh oh!

gewenyu99 commented Sep 22, 2026 •

edited

Loading

wizard-ci-bot Bot commented Sep 22, 2026 •

edited

Loading

johncwaters left a comment •

edited

Loading