@kici-dev/compiler 0.8.0 → 0.9.0

package/dist/llm-context/llms-getting-started.txt

@@ -845,7 +845,7 @@ Each guide is split into two parts. **Part 1** gets you a green run against your
 
 ## Bring a coding agent
 
-Workflows are TypeScript, so a coding agent can write them. KiCI ships its docs in a form an agent reads directly: point yours at [llms.txt](https://kici.dev/docs/llms.txt), or pipe a task bundle straight into its context with `kici docs llm` (`kici docs llm sdk` for the authoring API, `kici docs llm patterns` for recipes). An agent briefed that way can scaffold your first workflow, compile it, and read the failure when it breaks.
+Workflows are TypeScript, so a coding agent can write them. KiCI ships its docs in a form an agent reads directly: point yours at [llms.txt](https://kici.dev/llms.txt), or pipe a task bundle straight into its context with `kici docs llm` (`kici docs llm sdk` for the authoring API, `kici docs llm patterns` for recipes). An agent briefed that way can scaffold your first workflow, compile it, and read the failure when it breaks.
 
 If your agent finds that these docs promise something KiCI does not do, that is worth telling us — `kici feedback` prints how to report it, and [Reporting a discrepancy](https://docs.kici.dev/user/reporting-discrepancies/) is the full guide.
 
@@ -891,8 +891,10 @@ promised. You can report that.
 A human reader can use this page too. The rules below exist because the tracker
 is public and you are filing under someone else's GitHub identity.
 
-Run `kici feedback` to print this contract in your terminal, or
-`kici feedback --json` to read it as structured data.
+Run `kici feedback` to print this contract in your terminal,
+`kici feedback --json` to read it as structured data, or
+`kici feedback --draft <file.json> --open` to turn a draft into the prefilled form
+and open it.
 
 ## The rule, in one sentence
 
@@ -967,19 +969,20 @@ approval rule.
 
 ## What the report must carry
 
-Five things. A report missing any of them cannot be acted on:
+A report missing any of these fields cannot be acted on:
 
 | Field                                 | What it holds                                                                                                        |
 | ------------------------------------- | -------------------------------------------------------------------------------------------------------------------- |
 | What the docs or CLI advertise        | The exact claim, quoted, plus its source: a docs URL or the command whose `--help` says it.                          |
 | What actually happened                | The real output or behaviour, quoted, with any error text.                                                           |
 | Minimal reproduction, including setup | Every step from an empty directory: the setup commands, a minimal synthetic workflow, and the exact command you ran. |
-| Version and environment               | Output of `kici --version`, plus Node version and OS.                                                                |
+| KiCI version                          | Output of `kici --version`.                                                                                          |
+| Environment                           | Node version and OS.                                                                                                 |
 | Why this is a discrepancy             | One or two sentences ruling out the likely misreads — why the docs cannot be read to match what you observed.        |
 
 The last field is the one agents skip, and it is the one that makes a report
 usable. "The docs say `--foo` exists and it does not" is a claim. "`--foo` is
-documented at <url>, and `kici bar --help` on 0.1.16 lists no such flag" is a
+documented at <url>, and `kici bar --help` on the version I ran lists no such flag" is a
 finding someone can act on in one pass.
 
 Include setup steps whenever your reproduction needed any. A reproduction that
@@ -988,22 +991,36 @@ starts from state a maintainer cannot recreate is not a reproduction.
 ## Filing it
 
 The tracker is [kici-dev/kici-public](https://github.com/kici-dev/kici-public),
-and the form is **Agent report**:
+and the form is **Agent report** (`agent_report.yml`). It is a GitHub issue
+form, which only the browser can submit — `gh issue create` cannot fill it.
+So you draft, and the person files.
 
-```bash
-kici feedback --open
+Write the draft as JSON, one key per field of the form (`kici feedback --json`
+lists them under `draftFields`):
+
+```json
+{
+  "title": "kici compile --check writes a lock file",
+  "advertised": "https://docs.kici.dev/user/cli-reference/ says: `kici compile --check` type-checks without writing a lock file.",
+  "observed": ".kici/kici.lock.json was written; mtime changed.",
+  "reproduction": "$ mkdir /tmp/repro && cd /tmp/repro\n$ kici init\n$ kici compile --check",
+  "version": "0.8.0",
+  "environment": "Node 24.13.0, Ubuntu 24.04",
+  "justification": "The sentence names --check as the no-write mode; no reading of it allows a write."
+}
 ```
 
-Draft the complete body first. Show it to the person you are working with —
-the whole body, not a summary of it. File only after they say yes:
+Then build the prefilled form:
 
 ```bash
-gh issue create --repo kici-dev/kici-public --template agent_report.yml
+kici feedback --draft draft.json --open
 ```
 
-The form asks you to confirm the report was drafted by an agent and reviewed by
-a person before filing. Answer honestly; it is what lets maintainers triage
-agent-filed reports as a group.
+The command prints the form URL with every field filled in and, with
+`--open`, opens it. It reaches no network and files nothing. Show the person
+you are working with the whole draft — not a summary of it. They open the
+URL, read the prefilled form, tick the review boxes, and submit. File only
+after they say yes.
 
 If they say no, that is the end of it. Do not re-file the same finding in
 another form, and do not open it as a question instead.

package/dist/llm-context/llms-providers.txt

@@ -758,8 +758,8 @@ kici-admin org-settings global-workflows deny-add \
 ```
 
 See [Global workflows](https://docs.kici.dev/architecture/global-workflows/) for the
-policy model (`isWorkflowRepoAllowed` + `isSourceRepoAllowed` +
-`isElevatedAccessAllowed`) and the cross-provider dispatch contract.
+policy model (`isWorkflowRepoAllowed` + `isSourceRepoAllowed`) and the
+cross-provider dispatch contract.
 
 ## Routing-key collisions
 

package/dist/llm-context/llms-sdk-runtime.txt

@@ -1543,13 +1543,15 @@ const deploy = job('deploy', {
 
 Request a short-lived OIDC ID token for the current job, bound to an `audience`. The token is a signed JWT whose identity claims (`repository`, `ref`, `sha`, `kici_run_id`, `kici_job_id`) are derived by your orchestrator from the run context — a step cannot spoof them. Use it to authenticate the build to an external service that trusts the orchestrator's OIDC issuer (for example, when generating build provenance).
 
-The token also carries the event context a cloud trust policy needs to tell a fork pull request from a trusted push — `is_fork`, `head_repository`, `trust_tier`, `event_name`, and a pull-request-specific `sub`. See [ID-token claims and cloud trust policies](https://docs.kici.dev/user/provenance/#id-token-claims-and-cloud-trust-policies) for the full claim table and a worked AWS policy.
+The token also carries the event context a cloud trust policy needs to tell a fork pull request from a trusted push — `is_fork`, `head_repository`, `trust_tier`, `event_name`, and a pull-request-specific `sub`. See [ID-token claims and cloud trust policies](https://docs.kici.dev/user/oidc/#id-token-claims-and-cloud-trust-policies) for the full claim table and a worked AWS policy.
 
 ```typescript
 const publish = job('publish', {
   steps: [
     step('mint', async (ctx) => {
-      const { token, expiresIn } = await ctx.kici.oidc.token({ audience: 'sigstore' });
+      const minted = await ctx.kici.oidc.token({ audience: 'sigstore' });
+      if ('deferred' in minted) throw new Error(`ID token deferred: ${minted.code}`);
+      const { token, expiresIn } = minted;
       ctx.log.info(`Got an ID token valid for ${expiresIn}s`);
       // Hand `token` to a tool that exchanges it with the trusting service.
     }),
@@ -1559,6 +1561,7 @@ const publish = job('publish', {
 
 **Behavior:**
 
+- The result is the minted token, or `{ deferred: true, code }` (`unavailable` | `failed`) when the orchestrator could not mint one right now. A step that exchanges the token fails on `deferred`; `ctx.attestProvenance` handles it by freezing the statement for later fulfilment.
 - The token is short-lived (about 10 minutes) and scoped to the current run and job.
 - The returned token value is automatically masked in step logs.
 - The step never holds signing credentials — the orchestrator mints and signs the token on the step's behalf from its own run records.

package/dist/llm-context/llms-sdk.txt

@@ -1051,9 +1051,8 @@ The throw is a `ChangedFilesUnavailableError` (exported from `@kici-dev/sdk`, ca
 ### evaluateRules(rules, context, label, onRuleResult?)
 
 The agent calls this on your behalf. A workflow does not call it. It lives on
-`@kici-dev/sdk/internal`, outside semver, and stays exported from the root barrel as
-`@deprecated` until v1.0.0 — see [deprecations](https://docs.kici.dev/user/deprecations/). It is described here
-because its return shape is what a rule's outcome looks like in the run log.
+`@kici-dev/sdk/internal`, outside semver, and is not exported from the root barrel. It is
+described here because its return shape is what a rule's outcome looks like in the run log.
 
 Evaluate an array of rules sequentially with fail-fast behavior. Stops on the first failure.
 
@@ -1306,9 +1305,8 @@ isDynamicFunction(matrix); // true if async function
 ### Matrix expansion utilities
 
 The agent expands a matrix for you. A workflow does not call these. They live on
-`@kici-dev/sdk/internal`, outside semver, and stay exported from the root barrel as
-`@deprecated` until v1.0.0 — see [deprecations](https://docs.kici.dev/user/deprecations/). They are described
-here because they define the combinations a matrix job actually produces.
+`@kici-dev/sdk/internal`, outside semver, and are not exported from the root barrel. They are
+described here because they define the combinations a matrix job actually produces.
 
 ```typescript
 import { expandMatrix, applyIncludeExclude } from '@kici-dev/sdk/internal';
@@ -1905,7 +1903,7 @@ Create a commit status trigger. Returns a frozen `StatusTriggerConfig`.
 function status(config?: StatusConfigInput): StatusTriggerConfig;
 ```
 
-**Config options:** `contexts` (picomatch strings like 'ci/\*'), `states` (error/failure/pending/success), `description`
+**Config options:** `contexts` (glob patterns like 'ci/\*'), `states` (error/failure/pending/success), `description`
 
 ```typescript
 status(); // Any status
@@ -2531,11 +2529,8 @@ a `.result` proxy reads, build the step context your workflow body receives, eva
 rules, and expand its matrix. The agent drives all of it on your behalf.
 
 It is **not covered by semver** and may change shape in any release. Do not import it from
-a workflow. Everything a workflow author needs is on the root entry point above.
-
-Those same symbols are also still exported from the root barrel, marked `@deprecated`, so
-an older SDK in a repository keeps working. They are removed from the root at v1.0.0 — see
-[deprecations](https://docs.kici.dev/user/deprecations/).
+a workflow. Everything a workflow author needs is on the root entry point above, and none
+of the runtime contract is exported there.
 
 ## See also
 

package/dist/llm-context/llms.txt

@@ -1,6 +1,6 @@
 # KiCI
 
-> KiCI is a TypeScript-native CI/CD workflow engine. Workflows are defined in TypeScript (not YAML), compiled into a portable lock file, and executed by self-hosted agents. The docs below cover the SDK, the CLI, workflow patterns, and the runtime architecture an LLM coding agent needs in order to author and test KiCI workflows.
+> KiCI is a TypeScript-native CI/CD workflow engine. Workflows are defined in TypeScript (not YAML), compiled into a portable lock file, and executed by self-hosted agents. The docs below cover the SDK, the CLI, workflow patterns, and the runtime architecture an LLM coding agent needs to author and test KiCI workflows.
 
 The full markdown bundle of every page indexed here is available at https://docs.kici.dev/llms-full.txt.
 
@@ -10,16 +10,16 @@ Found a mismatch between what these docs advertise and what KiCI does? Read http
 
 Each bundle below is a self-contained markdown file for one authoring task. Fetch only the one your task needs instead of the full bundle:
 
-- [getting-started](https://docs.kici.dev/llms-getting-started.txt) (65 KB) — Adopt KiCI: why it exists, how workflows execute, migrating from GitHub Actions, installing the SDK, and writing/compiling/testing your first workflow
+- [getting-started](https://docs.kici.dev/llms-getting-started.txt) (66 KB) — Adopt KiCI: why it exists, how workflows execute, migrating from GitHub Actions, installing the SDK, and writing/compiling/testing your first workflow
 - [patterns](https://docs.kici.dev/llms-patterns.txt) (61 KB) — Copy-paste workflow recipes: triggers, conditionals, matrix, scheduling, integrations
 - [sdk](https://docs.kici.dev/llms-sdk.txt) (131 KB) — Core authoring API: workflow/job/step factories, triggers, rules, matrix, dynamic jobs, cross-job outputs
-- [sdk-runtime](https://docs.kici.dev/llms-sdk-runtime.txt) (115 KB) — Runtime and advanced authoring: step runtime context, event payloads, host fan-out, idempotent steps, caching, artifacts, waiting
+- [sdk-runtime](https://docs.kici.dev/llms-sdk-runtime.txt) (116 KB) — Runtime and advanced authoring: step runtime context, event payloads, host fan-out, idempotent steps, caching, artifacts, waiting
 - [cli](https://docs.kici.dev/llms-cli.txt) (97 KB) — Running the CLI locally: compile, test, run local, hooks, lock-file drift, common failures
 - [cli-remote](https://docs.kici.dev/llms-cli-remote.txt) (111 KB) — Auth, org and orchestrator selection, runs, approvals, notifications, diagnostics, and the MCP server a coding agent connects to
-- [features](https://docs.kici.dev/llms-features.txt) (97 KB) — Deployment contexts, scoped secrets, private registries, approvals, provenance, dashboard and account access
+- [features](https://docs.kici.dev/llms-features.txt) (112 KB) — Deployment contexts, scoped secrets, private registries, approvals, workload identity, provenance, dashboard and account access
 - [features-execution](https://docs.kici.dev/llms-features-execution.txt) (124 KB) — Concurrency, dynamic values, events, container jobs, environment variables, global workflows, idempotent steps
 - [providers](https://docs.kici.dev/llms-providers.txt) (35 KB) — Connecting sources: GitHub App, universal-git (Forgejo/Gitea/GitLab), local file://
-- [architecture](https://docs.kici.dev/llms-architecture.txt) (100 KB) — How the runtime works: three-tier relay model, data flows, configuration
+- [architecture](https://docs.kici.dev/llms-architecture.txt) (101 KB) — How the runtime works: three-tier relay model, data flows, configuration
 
 ## Getting started
 
@@ -86,6 +86,7 @@ Each bundle below is a self-contained markdown file for one authoring task. Fetc
 - [Approval gates](https://docs.kici.dev/user/approvals/): Pause a workflow for human sign-off at step, job, or workflow granularity with approval
 - [Contexts](https://docs.kici.dev/user/contexts/): Configure deployment contexts with variables, secrets, and protection rules
 - [Dashboard](https://docs.kici.dev/user/dashboard/): Web UI for monitoring workflow runs, managing sources, secrets, and organization settings.
+- [Workload identity with OIDC](https://docs.kici.dev/user/oidc/): Mint short-lived OIDC identity tokens in a step and exchange them with Cloudsmith, AWS, or any service that federates with your orchestrator
 - [Private npm registries](https://docs.kici.dev/user/private-registries/): Authenticate `npm install` against private registries (CodeArtifact, GitHub Packages, Verdaccio, …) from a workflow's `.kici/package.json`
 - [Build provenance and attestations](https://docs.kici.dev/user/provenance/): Generate and verify signed SLSA provenance for the artifacts your workflows build
 - [Secrets](https://docs.kici.dev/user/secrets/): How to access secrets in KiCI workflow steps

package/dist/local-plane/orchestrator-process.d.ts

@@ -26,15 +26,14 @@ export interface SpawnOrchestratorOptions {
      * generated by the plane. Enables the orchestrator's local OIDC mint +
      * dev-signed provenance (issuer `kici-local`) in independent mode. Present in
      * the independent boot; omitted (with `attach`) in the hybrid boot, which
-     * mints via the Platform relay instead.
+     * mints with the orchestrator's own signing key instead.
      */
     devIdentityKeyFile?: string;
     /**
      * When present, boot the orchestrator in HYBRID mode attached to the Platform
-     * relay instead of independent mode. OIDC + attestation then mint via the
-     * Platform (the local dev signer is never consulted — the orchestrator's
-     * `selectOidcMintRegistration` prefers the relay whenever a Platform
-     * connection exists).
+     * relay instead of independent mode. OIDC + attestation then mint with the
+     * orchestrator's own signing key under its own issuer, exactly as a deployed
+     * orchestrator does (the local dev signer is never consulted).
      */
     attach?: {
         /** Orchestrator KICI_PLATFORM_URL — a ws(s)://…/ws relay URL. */

package/dist/local-plane/orchestrator-process.js

@@ -77,7 +77,8 @@ function spawnOrchestratorProcess(databaseUrl, opts) {
 		modeEnv = {
 			KICI_MODE: "hybrid",
 			KICI_PLATFORM_URL: opts.attach.platformWsUrl,
-			KICI_PLATFORM_TOKEN: opts.attach.platformToken
+			KICI_PLATFORM_TOKEN: opts.attach.platformToken,
+			KICI_ORCHESTRATOR_PROVENANCE_ISSUER: `http://127.0.0.1:${port}`
 		};
 	} else {
 		entry = resolveStandaloneEntry();

package/dist/local-plane/plane-manager.js

@@ -19,8 +19,8 @@ import { generateKeyPairSync, randomBytes } from "node:crypto";
 */
 function currentBuildIdentity() {
 	return {
-		version: "0.8.0",
-		buildCommit: "57712173f"
+		version: "0.9.0",
+		buildCommit: "69765a7ee"
 	};
 }
 /**

package/dist/templates/agents-md.d.ts

@@ -4,5 +4,5 @@
  * picked up by Claude Code, Cursor, Aider, and other coding agents that scan
  * the working tree for an authoring context file.
  */
-export declare const agentsMdTemplate = "# KiCI workflow authoring guide\n\nThis project uses KiCI \u2014 a TypeScript-native CI/CD workflow engine \u2014 instead\nof YAML-based CI. Workflows live in `.kici/workflows/*.ts`, are compiled\ninto a portable lock file, and executed by self-hosted agents.\n\n## Where the API surface lives\n\n- Public SDK types: `node_modules/@kici-dev/sdk/dist/index.d.ts` \u2014 read this\n  for the canonical signatures of `workflow`, `job`, `step`, `pr`,\n  `push`, `schedule`, `rule`, `dynamicJob`, etc.\n- Bundled offline reference for coding agents: `kici docs llm` prints the\n  full markdown documentation bundle to stdout. `kici docs llm --index`\n  prints just the curated link index (llms.txt format).\n- Online docs:\n  - <https://kici.dev/docs/> \u2014 published docs site.\n  - <https://kici.dev/llms.txt> \u2014 curated index for LLM consumers.\n  - <https://kici.dev/llms-full.txt> \u2014 full markdown bundle.\n  - Key pages: `user/sdk-reference`, `user/workflow-patterns`,\n    `user/testing-guide`, `user/hooks`, `user/secrets`.\n\n## The five core patterns\n\n1. **Push trigger** \u2014 `on: push({ branches: 'main' })`. Pair with `paths`\n   to scope to subtrees.\n\n   ```ts\n   import { workflow, job, step, push } from '@kici-dev/sdk';\n\n   export default workflow('build', {\n     on: push({ branches: 'main' }),\n     jobs: [\n       job('build', {\n         runsOn: 'kici:os:linux',\n         steps: [step('install', async ({ $ }) => { await $`pnpm install`; })],\n       }),\n     ],\n   });\n   ```\n\n   `kici:os:linux` targets any agent reporting that OS \u2014 every agent\n   self-reports `kici:os:` / `kici:arch:` / `kici:host:`. Use a custom label\n   (e.g. `'gpu'`, `'prod-pool'`) to target a specific agent pool your scaler\n   defines.\n\n2. **PR + matrix** \u2014 `pr({ target: 'main' })` plus a matrix over node\n   versions. The matrix expands at dispatch time.\n\n   ```ts\n   import { workflow, job, step, pr } from '@kici-dev/sdk';\n\n   export default workflow('test-matrix', {\n     on: pr({ target: 'main' }),\n     jobs: [\n       job('test', {\n         runsOn: 'kici:os:linux',\n         matrix: { node: ['20', '22', '24'] },\n         steps: [\n           step('test', async ({ $, matrix }) => {\n             await $`echo testing on node ${matrix!.node}`;\n             await $`pnpm install`;\n             await $`pnpm test`;\n           }),\n         ],\n       }),\n     ],\n   });\n   ```\n\n3. **Lifecycle hooks** \u2014 `onFailure` / `onSuccess` / `onCancel` on a\n   job or workflow run after the main steps in their own scope.\n\n4. **Secrets** \u2014 declared scopes resolve at dispatch:\n\n   ```ts\n   step('deploy', async ({ $, secrets }) => {\n     await secrets.expose('DEPLOY_TOKEN');\n     await $`./scripts/deploy.sh`;\n   });\n   ```\n\n   Run `kici secrets list` to enumerate the contexts available for testing.\n\n5. **Dynamic jobs** \u2014 `dynamicJob` and `dynamicGroup` build the DAG at\n   runtime from a step's outputs. Don't try to compute job names at top level;\n   the lock file would be wrong.\n\n## Anti-patterns\n\n- **Do NOT write `.yml` / `.yaml` CI files** \u2014 KiCI replaces that entire\n  layer. There is no compatibility shim.\n- **Do NOT `import` from any `@kici-dev/*` package's `/dist/...`\n  subpath** \u2014 those are not part of the public API and break across versions.\n  Import from the package root.\n- **Do NOT `await` outside step bodies.** The top-level workflow file is\n  loaded by the compiler synchronously; async I/O at module scope means the\n  lock file emits before it resolves and the workflow appears empty.\n- **Do NOT mutate shared variables between jobs.** Each job runs in its own\n  agent process. Use `needs` + step outputs to thread values.\n- **Do NOT hand-edit `kici.lock.json`.** Regenerate it via `kici compile`.\n\n## Local commands a coding agent should run\n\n| Command                         | Purpose                                     |\n| ------------------------------- | ------------------------------------------- |\n| `pnpm kici compile --check`     | Validate workflow source without writing.   |\n| `pnpm kici preview pr:open --debug` | Preview which workflows match an event.   |\n| `pnpm kici run push --local`    | Execute a workflow locally (this machine as an ephemeral agent). |\n| `pnpm kici docs llm`            | Print the full LLM documentation bundle.    |\n| `pnpm kici docs llm --index`    | Print the curated link index.               |\n\nIf `pnpm kici` isn't in scripts, fall back to `npx kici`.\n\n## Loop\n\n1. Read the SDK types from `node_modules/@kici-dev/sdk/dist/index.d.ts`.\n2. Pipe `kici docs llm` into the agent's context if it doesn't already have\n   the full bundle.\n3. Edit a workflow under `.kici/workflows/`.\n4. Run `kici compile --check` (zero exit means valid).\n5. Run `kici preview <event>` to preview matching.\n6. Run `kici run <event> --local` to execute locally before pushing.\n";
+export declare const agentsMdTemplate = "# KiCI workflow authoring guide\n\nThis project uses KiCI \u2014 a TypeScript-native CI/CD workflow engine \u2014 instead\nof YAML-based CI. Workflows live in `.kici/workflows/*.ts`, are compiled\ninto a portable lock file, and executed by self-hosted agents.\n\n## Where the API surface lives\n\n- Public SDK types: `node_modules/@kici-dev/sdk/dist/index.d.ts` \u2014 read this\n  for the canonical signatures of `workflow`, `job`, `step`, `pr`,\n  `push`, `schedule`, `rule`, `dynamicJob`, etc.\n- Bundled offline reference for coding agents: `kici docs llm` prints the\n  curated link index (llms.txt format); `kici docs llm full` prints the\n  complete markdown documentation bundle; `kici docs llm <topic>` prints one\n  task bundle (`sdk`, `cli`, `patterns`, \u2026).\n- Online docs:\n  - <https://docs.kici.dev/> \u2014 published docs site.\n  - <https://kici.dev/llms.txt> \u2014 curated index for LLM consumers.\n  - <https://kici.dev/llms-full.txt> \u2014 full markdown bundle.\n  - Key pages: `user/sdk-reference`, `user/workflow-patterns`,\n    `user/testing-guide`, `user/hooks`, `user/secrets`.\n\n## The five core patterns\n\n1. **Push trigger** \u2014 `on: push({ branches: 'main' })`. Pair with `paths`\n   to scope to subtrees.\n\n   ```ts\n   import { workflow, job, step, push } from '@kici-dev/sdk';\n\n   export default workflow('build', {\n     on: push({ branches: 'main' }),\n     jobs: [\n       job('build', {\n         runsOn: 'kici:os:linux',\n         steps: [step('install', async ({ $ }) => { await $`pnpm install`; })],\n       }),\n     ],\n   });\n   ```\n\n   `kici:os:linux` targets any agent reporting that OS \u2014 every agent\n   self-reports `kici:os:` / `kici:arch:` / `kici:host:`. Use a custom label\n   (e.g. `'gpu'`, `'prod-pool'`) to target a specific agent pool your scaler\n   defines.\n\n2. **PR + matrix** \u2014 `pr({ target: 'main' })` plus a matrix over node\n   versions. The matrix expands at dispatch time.\n\n   ```ts\n   import { workflow, job, step, pr } from '@kici-dev/sdk';\n\n   export default workflow('test-matrix', {\n     on: pr({ target: 'main' }),\n     jobs: [\n       job('test', {\n         runsOn: 'kici:os:linux',\n         matrix: { node: ['20', '22', '24'] },\n         steps: [\n           step('test', async ({ $, matrix }) => {\n             await $`echo testing on node ${matrix!.node}`;\n             await $`pnpm install`;\n             await $`pnpm test`;\n           }),\n         ],\n       }),\n     ],\n   });\n   ```\n\n3. **Lifecycle hooks** \u2014 `onFailure` / `onSuccess` / `onCancel` on a\n   job or workflow run after the main steps in their own scope.\n\n4. **Secrets** \u2014 declared scopes resolve at dispatch:\n\n   ```ts\n   step('deploy', async ({ $, secrets }) => {\n     await secrets.expose('DEPLOY_TOKEN');\n     await $`./scripts/deploy.sh`;\n   });\n   ```\n\n   Run `kici secrets list` to enumerate the contexts available for testing.\n\n5. **Dynamic jobs** \u2014 `dynamicJob` and `dynamicGroup` build the DAG at\n   runtime from a step's outputs. Don't try to compute job names at top level;\n   the lock file would be wrong.\n\n## Anti-patterns\n\n- **Do NOT write `.yml` / `.yaml` CI files** \u2014 KiCI replaces that entire\n  layer. There is no compatibility shim.\n- **Do NOT `import` from any `@kici-dev/*` package's `/dist/...`\n  subpath** \u2014 those are not part of the public API and break across versions.\n  Import from the package root.\n- **Do NOT `await` outside step bodies.** The top-level workflow file is\n  loaded by the compiler synchronously; async I/O at module scope means the\n  lock file emits before it resolves and the workflow appears empty.\n- **Do NOT mutate shared variables between jobs.** Each job runs in its own\n  agent process. Use `needs` + step outputs to thread values.\n- **Do NOT hand-edit `kici.lock.json`.** Regenerate it via `kici compile`.\n\n## Local commands a coding agent should run\n\n| Command                         | Purpose                                     |\n| ------------------------------- | ------------------------------------------- |\n| `pnpm kici compile --check`     | Validate workflow source without writing.   |\n| `pnpm kici preview pr:open --debug` | Preview which workflows match an event.   |\n| `pnpm kici run push --local`    | Execute a workflow locally (this machine as an ephemeral agent). |\n| `pnpm kici docs llm`            | Print the curated link index (llms.txt).    |\n| `pnpm kici docs llm full`       | Print the full LLM documentation bundle.    |\n| `pnpm kici docs llm sdk`        | Print one task bundle (also `cli`, `patterns`, \u2026). |\n\nIf `pnpm kici` isn't in scripts, fall back to `npx kici`.\n\n## Loop\n\n1. Read the SDK types from `node_modules/@kici-dev/sdk/dist/index.d.ts`.\n2. Pipe `kici docs llm sdk` (or `kici docs llm full`) into the agent's context if\n   it doesn't already have the authoring reference.\n3. Edit a workflow under `.kici/workflows/`.\n4. Run `kici compile --check` (zero exit means valid).\n5. Run `kici preview <event>` to preview matching.\n6. Run `kici run <event> --local` to execute locally before pushing.\n";
 //# sourceMappingURL=agents-md.d.ts.map

package/dist/templates/agents-md.js

@@ -18,10 +18,11 @@ into a portable lock file, and executed by self-hosted agents.
   for the canonical signatures of \`workflow\`, \`job\`, \`step\`, \`pr\`,
   \`push\`, \`schedule\`, \`rule\`, \`dynamicJob\`, etc.
 - Bundled offline reference for coding agents: \`kici docs llm\` prints the
-  full markdown documentation bundle to stdout. \`kici docs llm --index\`
-  prints just the curated link index (llms.txt format).
+  curated link index (llms.txt format); \`kici docs llm full\` prints the
+  complete markdown documentation bundle; \`kici docs llm <topic>\` prints one
+  task bundle (\`sdk\`, \`cli\`, \`patterns\`, …).
 - Online docs:
-  - <https://kici.dev/docs/> — published docs site.
+  - <https://docs.kici.dev/> — published docs site.
   - <https://kici.dev/llms.txt> — curated index for LLM consumers.
   - <https://kici.dev/llms-full.txt> — full markdown bundle.
   - Key pages: \`user/sdk-reference\`, \`user/workflow-patterns\`,
@@ -114,16 +115,17 @@ into a portable lock file, and executed by self-hosted agents.
 | \`pnpm kici compile --check\`     | Validate workflow source without writing.   |
 | \`pnpm kici preview pr:open --debug\` | Preview which workflows match an event.   |
 | \`pnpm kici run push --local\`    | Execute a workflow locally (this machine as an ephemeral agent). |
-| \`pnpm kici docs llm\`            | Print the full LLM documentation bundle.    |
-| \`pnpm kici docs llm --index\`    | Print the curated link index.               |
+| \`pnpm kici docs llm\`            | Print the curated link index (llms.txt).    |
+| \`pnpm kici docs llm full\`       | Print the full LLM documentation bundle.    |
+| \`pnpm kici docs llm sdk\`        | Print one task bundle (also \`cli\`, \`patterns\`, …). |
 
 If \`pnpm kici\` isn't in scripts, fall back to \`npx kici\`.
 
 ## Loop
 
 1. Read the SDK types from \`node_modules/@kici-dev/sdk/dist/index.d.ts\`.
-2. Pipe \`kici docs llm\` into the agent's context if it doesn't already have
-   the full bundle.
+2. Pipe \`kici docs llm sdk\` (or \`kici docs llm full\`) into the agent's context if
+   it doesn't already have the authoring reference.
 3. Edit a workflow under \`.kici/workflows/\`.
 4. Run \`kici compile --check\` (zero exit means valid).
 5. Run \`kici preview <event>\` to preview matching.

package/dist/templates/package-json.js

@@ -1,6 +1,6 @@
 import "../rolldown-runtime-ClRpJifh.js";
 //#region src/templates/package-json.ts
-const sdkVersion = "0.8.0";
+const sdkVersion = "0.9.0";
 /**
 * The npm version spec the scaffold pins `@kici-dev/sdk` to.
 *

package/dist/templates/workflows/hello-world.ts

@@ -1,5 +1,5 @@
 // Hello World -- minimal push workflow
-// Docs: https://kici.dev/docs/sdk-reference
+// Docs: https://docs.kici.dev/user/sdk-reference/
 
 import { workflow, job, step, push } from '@kici-dev/sdk';
 

package/dist/templates/workflows/pr-checks.ts

@@ -1,6 +1,6 @@
 // PR Checks -- workflow with rules, dependencies, and multiple jobs
-// Docs: https://kici.dev/docs/sdk-reference
-// Patterns: https://kici.dev/docs/workflow-patterns
+// Docs: https://docs.kici.dev/user/sdk-reference/
+// Patterns: https://docs.kici.dev/user/workflow-patterns/
 
 import { workflow, job, step, pr, rule, skip, isEventType } from '@kici-dev/sdk';
 

package/dist/test-runner/job-executor.js

@@ -12,13 +12,14 @@ import { setJobOutputsMap, setStepOutputsMap, setStepRefMap } from "@kici-dev/sd
 /**
 * Resolve SDK output setter functions from the workflow's module instance.
 * This ensures the output maps are set on the same SDK module that the
-* workflow code uses for .result proxy resolution and ctx.outputsOf().
+* workflow code uses for .result proxy resolution and ctx.outputsOf(). The
+* setters live only on the `@kici-dev/sdk/internal` subpath.
 *
 * Falls back to the compiler's own SDK import if dynamic resolution fails.
 */
 async function resolveSdkSetters(kiciDir) {
 	if (kiciDir) try {
-		const sdkPath = path.join(kiciDir, "node_modules", "@kici-dev", "sdk", "dist", "index.js");
+		const sdkPath = path.join(kiciDir, "node_modules", "@kici-dev", "sdk", "dist", "internal.js");
 		const sdk = await import(pathToFileURL(sdkPath).href);
 		return {
 			setStepOutputsMap: sdk.setStepOutputsMap,

package/dist/types.d.ts

@@ -8,7 +8,7 @@
  * v6 replaces job-level contexts with environment/env/concurrencyGroup.
  * v7 adds hook flags, step rules, gracePeriod, and workflow concurrency config.
  * v8 adds runsOn polymorphic type (string | string[] | selector) and excludeLabels.
- * v11 adds LockInlineValue type for pure function inline evaluation.
+ * v11 added an inline-expression value shape for dynamic fields, which is no longer parsed.
  * v15 adds per-job init config(s).
  * v17 widens per-job init to typed presets ('mise' / { mise }) and 'auto' detection.
  */
@@ -394,29 +394,6 @@ export interface LockParallelStep {
 export type LockStepEntry = LockStep | LockParallelStep;
 /** Type guard distinguishing a parallel group from an ordinary lock step. */
 export declare function isLockParallelStep(entry: LockStepEntry): entry is LockParallelStep;
-/**
- * Serialized inline expression for a dynamic env/context/concurrencyGroup
- * field, shaped as `{ _type: 'inline', expression: '(event) => ...' }`
- * alongside the existing 'static' and 'dynamic' discriminants.
- *
- * @deprecated Schema v11 inline expressions are no longer evaluated in the
- * orchestrator. Dynamic env/context/concurrencyGroup fields are resolved on the
- * eval agent's init-runner. The compiler no longer emits this type; readers keep
- * recognizing it only to defer an old lock's field to the init round. Removed at
- * the next major (v1.0.0).
- */
-export interface LockInlineValue {
-    readonly _type: 'inline';
-    readonly expression: string;
-}
-/**
- * Type guard for inline expression values.
- *
- * @deprecated See {@link LockInlineValue}. Retained only so a reader can
- * recognize an old lock's inline field and defer it to the eval agent's
- * init-runner. Removed at the next major (v1.0.0).
- */
-export declare function isLockInlineValue(value: unknown): value is LockInlineValue;
 /**
  * Static job in lock file.
  * Contains all orchestrator-readable information for scheduling.
@@ -500,21 +477,21 @@ export interface LockJob {
         };
     };
     /**
-     * Bound contexts in merge order. Each entry is a static name or inline
-     * expression (pure function); `dynamic` is set when it is a function resolved at
-     * two-phase eval. Later entries override earlier ones on name collisions.
+     * Bound contexts in merge order. Each entry is a static name; `dynamic` is set
+     * when it is a function resolved on the eval agent's init-runner. Later entries
+     * override earlier ones on name collisions.
      */
     readonly contexts?: ReadonlyArray<{
-        value: string | LockInlineValue;
+        value: string;
         dynamic: boolean;
     }>;
-    /** Static environment variables or inline expression (pure function). */
-    readonly env?: Record<string, string> | LockInlineValue;
-    /** When true, env is dynamic (function) -- resolved at orchestrator two-phase eval or inline. */
+    /** Static environment variables. */
+    readonly env?: Record<string, string>;
+    /** When true, env is dynamic (function) -- resolved on the eval agent's init-runner. */
     readonly dynamicEnv?: boolean;
-    /** Concurrency group name (static string) or inline expression (pure function). */
-    readonly concurrencyGroup?: string | LockInlineValue;
-    /** When true, concurrencyGroup is dynamic (function) -- resolved at orchestrator two-phase eval or inline. */
+    /** Concurrency group name (static string). */
+    readonly concurrencyGroup?: string;
+    /** When true, concurrencyGroup is dynamic (function) -- resolved on the eval agent's init-runner. */
     readonly dynamicConcurrencyGroup?: boolean;
     /** Whether this job has an onCancel hook. */
     readonly hasOnCancel?: boolean;
@@ -658,7 +635,7 @@ export interface LockWorkflow {
  * v6 replaces job-level contexts with environment/env/concurrencyGroup.
  * v7 adds hook flags, step rules, gracePeriod, and workflow concurrency config.
  * v8 adds runsOn polymorphic type (string | string[] | selector) and excludeLabels.
- * v11 adds LockInlineValue type for pure function inline evaluation.
+ * v11 added an inline-expression value shape for dynamic fields, which is no longer parsed.
  * v13 adds job-level and workflow-level timeout.
  * v34 adds LockWorkflow.hasFilter (workflow-level pre-dispatch filter predicate).
  */

package/dist/types.js

@@ -11,7 +11,7 @@ import { BREAKING_FLOOR as BREAKING_FLOOR$1, SCHEMA_VERSION as SCHEMA_VERSION$1
 * v6 replaces job-level contexts with environment/env/concurrencyGroup.
 * v7 adds hook flags, step rules, gracePeriod, and workflow concurrency config.
 * v8 adds runsOn polymorphic type (string | string[] | selector) and excludeLabels.
-* v11 adds LockInlineValue type for pure function inline evaluation.
+* v11 added an inline-expression value shape for dynamic fields, which is no longer parsed.
 * v15 adds per-job init config(s).
 * v17 widens per-job init to typed presets ('mise' / { mise }) and 'auto' detection.
 */
@@ -23,16 +23,6 @@ const BREAKING_FLOOR = BREAKING_FLOOR$1;
 function isLockParallelStep(entry) {
 	return entry.kind === "parallel";
 }
-/**
-* Type guard for inline expression values.
-*
-* @deprecated See {@link LockInlineValue}. Retained only so a reader can
-* recognize an old lock's inline field and defer it to the eval agent's
-* init-runner. Removed at the next major (v1.0.0).
-*/
-function isLockInlineValue(value) {
-	return typeof value === "object" && value !== null && value._type === "inline";
-}
 /** Type guard for static jobs */
 function isLockStaticJob(job) {
 	return job._type === "static";
@@ -42,6 +32,6 @@ function isLockDynamicJobFn(job) {
 	return job._type === "dynamic";
 }
 //#endregion
-export { BREAKING_FLOOR, SCHEMA_VERSION, isLockDynamicJobFn, isLockInlineValue, isLockParallelStep, isLockStaticJob };
+export { BREAKING_FLOOR, SCHEMA_VERSION, isLockDynamicJobFn, isLockParallelStep, isLockStaticJob };
 
 //# sourceMappingURL=types.js.map

package/dist/types.test-d.d.ts

@@ -0,0 +1,2 @@
+export {};
+//# sourceMappingURL=types.test-d.d.ts.map

package/dist/types.test-d.js

@@ -0,0 +1,63 @@
+import "./rolldown-runtime-ClRpJifh.js";
+import { describe, expectTypeOf, it } from "vitest";
+//#region src/types.test-d.ts
+/**
+* The lock-file dynamic fields no longer admit the schema-v11 inline
+* expression `{ _type: 'inline', expression }`. That object is itself a
+* `Record<string, string>`, so the `env` axis is pinned by type identity rather
+* than by an assignment the compiler would accept either way; the two string
+* fields refuse the object outright.
+*/
+describe("LockJob dynamic fields", () => {
+	it("refuses an inline expression as the concurrency group", () => {
+		expectTypeOf({
+			_type: "static",
+			name: "build",
+			steps: [],
+			needs: [],
+			concurrencyGroup: {
+				_type: "inline",
+				expression: "() => \"deploy\""
+			}
+		}).toBeObject();
+	});
+	it("refuses an inline expression as a context name", () => {
+		expectTypeOf({
+			_type: "static",
+			name: "build",
+			steps: [],
+			needs: [],
+			contexts: [{
+				value: {
+					_type: "inline",
+					expression: "() => \"prod\""
+				},
+				dynamic: false
+			}]
+		}).toBeObject();
+	});
+	it("types env as exactly a plain record", () => {
+		expectTypeOf().toEqualTypeOf();
+	});
+	it("accepts the static shapes", () => {
+		const job = {
+			_type: "static",
+			name: "build",
+			steps: [],
+			needs: [],
+			env: { NODE_ENV: "test" },
+			dynamicEnv: true,
+			concurrencyGroup: "deploy",
+			contexts: [{
+				value: "production",
+				dynamic: false
+			}]
+		};
+		expectTypeOf(job.env).toEqualTypeOf();
+		expectTypeOf(job.concurrencyGroup).toEqualTypeOf();
+	});
+});
+//#endregion
+export {};
+
+//# sourceMappingURL=types.test-d.js.map

package/dist/workflows/hello-world.ts

@@ -1,5 +1,5 @@
 // Hello World -- minimal push workflow
-// Docs: https://kici.dev/docs/sdk-reference
+// Docs: https://docs.kici.dev/user/sdk-reference/
 
 import { workflow, job, step, push } from '@kici-dev/sdk';
 

package/dist/workflows/pr-checks.ts

@@ -1,6 +1,6 @@
 // PR Checks -- workflow with rules, dependencies, and multiple jobs
-// Docs: https://kici.dev/docs/sdk-reference
-// Patterns: https://kici.dev/docs/workflow-patterns
+// Docs: https://docs.kici.dev/user/sdk-reference/
+// Patterns: https://docs.kici.dev/user/workflow-patterns/
 
 import { workflow, job, step, pr, rule, skip, isEventType } from '@kici-dev/sdk';