akm-cli 0.8.0-rc2 → 0.8.0

package/dist/workflows/runs.js

@@ -1,3 +1,6 @@
+// This Source Code Form is subject to the terms of the Mozilla Public
+// License, v. 2.0. If a copy of the MPL was not distributed with this
+// file, You can obtain one at https://mozilla.org/MPL/2.0/.
 import { randomUUID } from "node:crypto";
 import fs from "node:fs";
 import { parseAssetRef } from "../core/asset-ref";
@@ -22,7 +25,7 @@ async function withWorkflowDb(fn) {
         closeWorkflowDatabase(db);
     }
 }
-export async function startWorkflowRun(ref, params = {}) {
+export async function startWorkflowRun(ref, params = {}, options) {
     const asset = await loadWorkflowAsset(ref);
     return withWorkflowDb(async (db) => {
         const now = new Date().toISOString();
@@ -30,6 +33,20 @@ export async function startWorkflowRun(ref, params = {}) {
         const scopeKey = getCurrentWorkflowScopeKey();
         const currentStepId = asset.steps[0]?.id ?? null;
         const workflowEntryId = resolveWorkflowEntryId(asset.sourcePath, asset.ref);
+        // Concurrency guard (#485): if an active run already exists in this
+        // (workflow_ref, scope_key) pair, refuse to create a parallel run unless
+        // `force: true` is set. Previously every call inserted unconditionally,
+        // so two terminals running `akm workflow start <ref>` left two runs
+        // racing; `akm workflow next` then non-deterministically picked one.
+        if (!options?.force) {
+            const existing = db
+                .prepare("SELECT id, current_step_id FROM workflow_runs WHERE workflow_ref = ? AND scope_key = ? AND status = 'active' ORDER BY updated_at DESC LIMIT 1")
+                .get(asset.ref, scopeKey);
+            if (existing) {
+                throw new UsageError(`Workflow ${asset.ref} already has an active run in this scope (id=${existing.id}, step=${existing.current_step_id ?? "—"}). ` +
+                    `Use 'akm workflow next ${asset.ref}' to resume it, 'akm workflow abandon ${existing.id}' to give up on it, or pass --force to start a parallel run.`, "RESOURCE_ALREADY_EXISTS");
+            }
+        }
         db.transaction(() => {
             db.prepare(`INSERT INTO workflow_runs (
           id, workflow_ref, scope_key, workflow_entry_id, workflow_title, status, params_json, current_step_id, created_at, updated_at

package/dist/workflows/schema.js

@@ -1,3 +1,6 @@
+// This Source Code Form is subject to the terms of the Mozilla Public
+// License, v. 2.0. If a copy of the MPL was not distributed with this
+// file, You can obtain one at https://mozilla.org/MPL/2.0/.
 /**
  * Validated JSON shape for a workflow asset.
  *

package/dist/workflows/scope-key.js

@@ -1,3 +1,6 @@
+// This Source Code Form is subject to the terms of the Mozilla Public
+// License, v. 2.0. If a copy of the MPL was not distributed with this
+// file, You can obtain one at https://mozilla.org/MPL/2.0/.
 import { createHash } from "node:crypto";
 import fs from "node:fs";
 import path from "node:path";

package/dist/workflows/validator.js

@@ -1,12 +1,8 @@
-/**
- * Cross-cutting semantic checks over an assembled WorkflowDocument draft.
- *
- * The parser handles per-line shape checks; this module runs rules that need
- * the whole document or the raw frontmatter at once: duplicate step IDs,
- * step-id format, and the frontmatter key whitelist.
- */
+// This Source Code Form is subject to the terms of the Mozilla Public
+// License, v. 2.0. If a copy of the MPL was not distributed with this
+// file, You can obtain one at https://mozilla.org/MPL/2.0/.
 const STEP_ID_REGEX = /^[A-Za-z0-9][A-Za-z0-9._-]*$/;
-const ALLOWED_FRONTMATTER_KEYS = new Set(["description", "tags", "params", "name", "updated"]);
+const ALLOWED_FRONTMATTER_KEYS = new Set(["description", "tags", "params", "name", "updated", "when_to_use"]);
 export function runSemanticChecks(draft, frontmatterData, frontmatterEndLine, errors) {
     checkFrontmatterKeys(frontmatterData, frontmatterEndLine, errors);
     checkStepIdFormat(draft, errors);
@@ -18,7 +14,7 @@ function checkFrontmatterKeys(data, fmEndLine, errors) {
             continue;
         errors.push({
             line: fmEndLine,
-            message: `Workflow frontmatter "${key}" is not supported. Use only: description, tags, params.`,
+            message: `Workflow frontmatter "${key}" is not supported. Use only: description, tags, params, when_to_use.`,
         });
     }
 }

package/docs/README.md

@@ -10,8 +10,9 @@
 
 ## Upgrading
 
+- [Roadmap](roadmap.md) -- High-level focus for the 0.9 and 1.0 releases
 - [v1 migration guide](migration/v1.md) -- The path from 0.x to v1.0, including the `.stash.json` removal scheduled for v0.8.0
-- [Release notes (latest: 0.7.0)](migration/release-notes/0.7.0.md) -- Per-release notes drop into `migration/release-notes/`, including current pre-release removals
+- [Release notes (latest: 0.8.0)](migration/release-notes/0.8.0.md) -- Per-release notes drop into `migration/release-notes/`, including current pre-release removals
 - [v0.5 → v0.6 migration guide](migration/v0.5-to-v0.6.md) -- Every breaking change with before/after code, publisher checklist, and troubleshooting
 
 ## Reference
@@ -28,6 +29,10 @@
 - [itlackey/akm-plugins](https://github.com/itlackey/akm-plugins) -- optional integrations for tools like OpenCode
 - [itlackey/akm-bench](https://github.com/itlackey/akm-bench) -- the standalone benchmark and evaluation repo for akm
 
+## Operations
+
+- Analyzing `akm improve` runs -- use [`akm health`](../src/commands/health.ts) (0.8.0+): `--since`, `--detail per-run`, `--window-compare`, `--windows`. See [health-command-enhancements.md](technical/health-command-enhancements.md).
+
 ## Internals
 
 - [Search](technical/search.md) -- Hybrid search architecture and scoring
@@ -41,7 +46,7 @@
 - [Ref Format](technical/ref.md) -- Wire format for asset references
 - [Test Coverage Guide](technical/test-coverage-guide.md) -- High-value testing areas
 - [Core Principles](technical/akm-core-principles.md) -- Design principles and constraints
-- [akm-bench](technical/benchmark.md) -- Search-quality benchmark suite
+- `technical/benchmark.md` (planned) -- Search-quality benchmark suite
 
 ## Posts
 

package/docs/data-and-telemetry.md

@@ -0,0 +1,225 @@
+# AKM Data & Telemetry
+
+AKM stores data locally on your machine. **It has no remote telemetry.** No data is sent to Anthropic, the AKM project, or any third party. This document describes exactly what AKM reads and writes on your machine.
+
+## No Remote Data Collection
+
+AKM does not:
+- Send usage data, events, or crash reports to any server
+- Contact any AKM-operated endpoint at runtime (only your own configured LLM/embedding endpoints)
+- Include any analytics SDK or beacon
+- Collect email, name, or any personally-identifying information
+
+The only network requests AKM makes are:
+1. Fetching registry metadata and stash packages from sources you explicitly configure (GitHub, npm, websites)
+2. Calls to your configured LLM/embedding endpoint (if you enabled those features)
+3. `akm upgrade` — fetches the latest release from GitHub releases
+4. `akm setup` — a single DNS lookup for `github.com` to decide whether to skip network-dependent steps (Ollama detection, remote embedding probes) when offline. No HTTP request is made by this probe; if it succeeds, akm proceeds with the network-dependent steps you already configured.
+
+---
+
+## Local On-Disk Surface
+
+AKM writes to these locations on your machine. All paths follow [XDG Base Directory](https://specifications.freedesktop.org/basedir-spec/latest/) conventions on Linux/macOS and Windows conventions on Windows.
+
+### Config Directory (`$XDG_CONFIG_HOME/akm` or `~/.config/akm/`)
+
+| Path | Contents | Safe to delete? |
+|---|---|---|
+| `config.json` | Your AKM configuration: LLM endpoints, stash paths, feature flags, profiles | **No** — deleting resets all settings |
+
+Override: set `AKM_CONFIG_DIR` or `XDG_CONFIG_HOME`.
+
+### Data Directory (`$XDG_DATA_HOME/akm` or `~/.local/share/akm/`)
+
+| Path | Contents | Safe to delete? |
+|---|---|---|
+| `index.db` | Search index for all your stash assets (FTS5 + metadata) | Yes — rebuilds via `akm index --full` |
+| `workflow.db` | Workflow run state and history | Caution — deletes run history |
+| `state.db` | Events, proposals, task history, improve run results | **No** — deletes event log, proposal queue, improve history |
+| `akm.lock` | Inter-process write lock | Yes — recreated automatically |
+| `akm.lock.lck` | Lock write sentinel | Yes — recreated automatically |
+
+Override: set `AKM_DATA_DIR` or `XDG_DATA_HOME`.
+
+### Cache Directory (`$XDG_CACHE_HOME/akm` or `~/.cache/akm/`)
+
+Everything in the cache is regenerable. It is safe to delete the entire cache directory; AKM will recreate what it needs on next use.
+
+| Path | Contents | Safe to delete? |
+|---|---|---|
+| `config-backups/config-<timestamp>.json` | Pre-migration config snapshots (5 retained) | Yes |
+| `config-backups/config.latest.json` | Latest backup alias | Yes |
+| `registry/` | Downloaded registry tarballs (stash packages from npm, GitHub, etc.) | Yes — re-downloaded on next `akm add` or `akm update` |
+| `registry-index/` | Legacy per-URL JSON cache (v0.7 artifact) | Yes — fully replaced by `index.db` in 0.8.0 |
+| `semantic-status.json` | Semantic index build status marker | Yes |
+| `bin/` | Downloaded AKM binary cache (used by `akm upgrade`) | Yes |
+| `tasks/logs/` | Scheduled task log files | Yes — ephemeral logs |
+| `tasks/history/` | Legacy task history JSONL (v0.7 migration artifact) | Yes |
+
+Override: set `AKM_CACHE_DIR` or `XDG_CACHE_HOME`.
+
+### Stash Directory (`~/akm/` by default, or user-configured)
+
+| Path | Contents | Safe to delete? |
+|---|---|---|
+| `<stash>/` | All your asset files: agents, skills, commands, knowledge, workflows, memories, vaults, wikis, lessons | **No** — this is YOUR data |
+| `<stash>/.akm/` | Hidden AKM metadata (v0.7 proposals, legacy runs) | Caution — check for pending proposals first |
+
+Override: set `AKM_STASH_DIR` (or configure `stashDir` in `config.json`).
+
+---
+
+## What Is Stored in `state.db`
+
+`state.db` holds three categories of non-regenerable data:
+
+### 1. Events Table
+
+An append-only log of every mutating action you perform with AKM. Events are stored locally for self-improvement (the improve loop uses them to surface usage patterns) and for inspection via `akm events`.
+
+**What is recorded:**
+- `event_type` — what action was taken (see full list below)
+- `ts` — ISO-8601 UTC timestamp
+- `ref` — the asset ref affected (e.g. `skill:code-review`), if applicable
+- `metadata` — structured payload specific to the event type (e.g. query text for `search`, score for `feedback`)
+
+**What is NOT recorded:**
+- File contents
+- LLM prompts or responses
+- API keys or secrets (config is not stored in events)
+- Personal information
+
+**Retention:** Events older than 90 days are purged automatically when `akm improve` runs its maintenance pass. Purge is controlled by `purgeOldEvents()` with a 90-day default.
+
+**Full event type list:**
+
+| Event type | When emitted | Key metadata fields |
+|---|---|---|
+| `add` | `akm add <source>` | `ref`, `provider` |
+| `remove` | `akm remove <source>` | `ref` |
+| `update` | `akm update [source]` | `ref` |
+| `remember` | `akm remember <text>` | `ref` |
+| `import` | `akm import <file>` | `ref` |
+| `save` | `akm save <ref>` | `ref` |
+| `feedback` | `akm feedback <ref>` | `signal` (positive/negative) |
+| `search` | `akm search <query>` | `query`, `source`, `signal` |
+| `curate` | `akm curate <prompt>` | `query`, `source` |
+| `show` | `akm show <ref>` | `ref`, `type`, `name` |
+| `select` | `akm show` after a search returning the same ref | `ref`, `entryId` |
+| `promoted` | `akm proposal accept <id>` | `ref` |
+| `rejected` | `akm proposal reject <id>` | `ref` |
+| `reflect_invoked` | Start of reflect phase in `akm improve` | `ref`, profile |
+| `reflect_completed` | Reflect phase produced a proposal | `ref` |
+| `improve_reflect_outcome` | Per-asset reflect result | `ref`, `ok`, `durationMs`, `reason` |
+| `propose_invoked` | `akm propose` | `ref` |
+| `distill_invoked` | `akm distill` | |
+| `improve_skipped` | Asset skipped by cooldown or budget | `ref`, `reason` |
+| `improve_completed` | `akm improve` run finished | run stats |
+| `improve_failed` | `akm improve` run errored | error |
+| `improve_lock_recovered` | Stale lock cleared at startup | |
+| `proposal_orphan_purge` | Stale proposals pruned | `checked`, `rejected` |
+| `proposal_creation_rejected` | `createProposal()` validation failed | `ref`, `reason`, `source` |
+| `proposal_expired` | Proposal expired | `ref` |
+| `events_purged` | Old events deleted by maintenance | `purgedCount`, `retentionDays` |
+| `workflow_started` | `akm workflow start <ref>` | `ref`, `runId` |
+| `workflow_step_completed` | `akm workflow next` | `ref`, `runId`, `stepId` |
+| `workflow_finished` | `akm workflow complete` | `ref`, `runId` |
+| `schema_repair_invoked` | `akm lint --repair` triggered schema repair | `ref` |
+| `archive_cleanup` | Archive cleanup during consolidation | |
+
+### 2. Proposals Table
+
+The proposal queue: pending, accepted, and rejected improvement proposals for your stash assets. Generated by `akm reflect` and the improve loop.
+
+Contents:
+- Proposal UUID (primary key)
+- Target asset ref
+- Status (pending/accepted/rejected)
+- Source (which process generated it — e.g. `reflect`, `distill`)
+- Full proposal content (Markdown text)
+- Created/updated timestamps
+
+### 3. Task History Table
+
+A record of scheduled task runs (from `akm tasks`):
+- Task ID, status, start/end times
+- Log file path (the log content stays in `$CACHE/tasks/logs/`)
+
+---
+
+## How to Inspect and Clear Local Data
+
+### Inspect events
+
+```sh
+# List recent events
+akm events list
+
+# Stream live events (tail)
+akm events tail
+
+# Filter by type
+akm events list --type search --limit 20
+
+# Filter by asset ref
+akm events list --ref skill:code-review
+```
+
+### Inspect proposals
+
+```sh
+# List pending proposals
+akm proposal list
+
+# Show a specific proposal
+akm proposal show <id>
+```
+
+### Clear specific data
+
+```sh
+# Delete the search index (safe — rebuilds with akm index --full)
+rm ~/.local/share/akm/index.db
+
+# Delete all cached registry downloads
+rm -rf ~/.cache/akm/registry/
+
+# Delete config backups
+rm -rf ~/.cache/akm/config-backups/
+
+# Delete the events log from state.db (non-reversible)
+# There is no akm CLI command to do this directly in 0.8.0.
+# Use SQLite directly:
+sqlite3 ~/.local/share/akm/state.db "DELETE FROM events;"
+
+# Delete all proposals
+sqlite3 ~/.local/share/akm/state.db "DELETE FROM proposals;"
+```
+
+### Start completely fresh (nuclear reset)
+
+```sh
+rm -f ~/.config/akm/config.json
+rm -rf ~/.local/share/akm/
+rm -rf ~/.cache/akm/
+# Your stash files in ~/akm/ are NOT touched by the above.
+```
+
+---
+
+## Environment Variable Overrides
+
+You can redirect any AKM directory to a custom path:
+
+| Variable | Overrides |
+|---|---|
+| `AKM_CONFIG_DIR` | Config directory (`~/.config/akm/`) |
+| `AKM_DATA_DIR` | Data directory (`~/.local/share/akm/`) |
+| `AKM_STATE_DIR` | State directory (`~/.local/state/akm/`) |
+| `AKM_CACHE_DIR` | Cache directory (`~/.cache/akm/`) |
+| `AKM_STASH_DIR` | Default stash directory (`~/akm/`) |
+| `XDG_CONFIG_HOME` | XDG base — akm appends `/akm` |
+| `XDG_DATA_HOME` | XDG base — akm appends `/akm` |
+| `XDG_STATE_HOME` | XDG base — akm appends `/akm` |
+| `XDG_CACHE_HOME` | XDG base — akm appends `/akm` |

package/docs/migration/release-notes/0.7.5.md

@@ -15,6 +15,6 @@ No manual migration is required for most users, but note these operator-facing d
 
 1. If you were relying on ref-based workflow runs being globally shared across repos or directories, that behavior is gone. Use direct run IDs when you intentionally need to resume a specific run created elsewhere.
 2. Existing older workflow runs created before the new workspace scope key was introduced are still accessible by direct run ID, but ref-based commands now prefer runs created in the current scope.
-3. `akm help migrate latest` now resolves to 0.7.5 and uses the published `.github/CHANGELOG.md` path.
+3. `akm help migrate latest` now resolves to 0.7.5 and uses the published `CHANGELOG.md` path.
 
-Full changelog: https://github.com/itlackey/akm/blob/main/.github/CHANGELOG.md
+Full changelog: https://github.com/itlackey/akm/blob/main/CHANGELOG.md

package/docs/migration/release-notes/0.8.0.md

@@ -9,6 +9,11 @@ Key operator-facing changes:
 - Plain `akm index` now owns metadata enhancement only.
 - Memory inference and graph extraction now run from `akm improve` after
   consolidation, not from `akm index`.
+- Graph schema redesign (DB_VERSION 12 → 13): `graph_files` re-keyed on
+  `entry_id` with FK cascade to `entries(id)`, child tables follow.
+  Upgrade drops the graph tables and repopulates them on the next
+  `akm improve` cycle; non-graph tables rebuild automatically. Run
+  `akm improve` once after the storage migration to repopulate graph data.
 - Manual QA and sandbox guidance now require isolating `AKM_DATA_DIR` in
   addition to `HOME`, XDG dirs, and `AKM_STASH_DIR`.
 - `akm health` is available for post-upgrade runtime checks against `state.db`,
@@ -18,11 +23,11 @@ Primary public command family for 0.8.0:
 
 - `akm improve [<type>|<ref>]`
 - `akm propose <type> <name> (--task "..." | --file <path>)`
-- `akm proposals`
-- `akm show proposal <id>`
-- `akm diff proposal <id>`
-- `akm accept <id>`
-- `akm reject <id> --reason "..."`
+- `akm proposal list`
+- `akm proposal show <id>`
+- `akm proposal diff <id>`
+- `akm proposal accept <id>`
+- `akm proposal reject <id> --reason "..."`
 
 Proposal-queue and improvement workflows are consolidated around this family.
 Update scripts, prompts, docs, and agent instructions before upgrading.
@@ -36,6 +41,53 @@ Behavior changes worth noting:
   `wikis/<name>/raw/`.
 - `akm health --since 24h` provides a fast operator check for state-db
   round-trip health, task-log backing, and recent improve outcomes.
+- The `akm improve` lock is now properly released even when `openStateDatabase()`
+  or budget-timer setup throws, preventing lock leaks (up to 4 hours) on error
+  paths.
+
+New in the final 0.8.0 RC cycle:
+
+### End-of-Run Git Sync (`akm improve --sync`)
+
+`akm improve` now auto-commits the git-backed primary stash at the end of every
+non-dry-run pass. Detection is based on a `.git` directory — no remote required.
+
+- `default` and `thorough` profiles: sync + push enabled by default
+- `quick` and `memory-focus` profiles: sync disabled (lightweight passes)
+- `--sync` / `--no-sync`: force on or off for a single run
+- `--push` / `--no-push`: push or commit-only override for a single run
+
+The commit message defaults to `akm improve auto-sync` and supports `{token}`
+placeholders: `{timestamp}`, `{date}`, `{time}`, `{scope}`, `{refs}`,
+`{accepted}`. Configure via `profiles.improve.<name>.sync.message` in config.
+
+The result envelope includes a `sync` field: `{ committed, pushed, skipped,
+reason? }`. A `stash_synced` event is emitted to `state.db`. Sync failures are
+non-fatal — they never fail a successful improve run.
+
+### Consolidate Reliability
+
+- Pre-flight stale-DB filter: consolidate now filters out entries that have
+  become stale in the DB before making any LLM calls, eliminating wasted tokens
+  on already-processed content.
+- Defense-in-depth merge guards: three layers of guards prevent incorrect merges,
+  improving consolidation correctness on edge cases.
+
+### Proposal Drain Refinements
+
+- `DrainDeferReason` now has two variants: `"mid-band"` and `"possible-dup"`.
+  The `"possible-contradiction"` variant has been removed.
+- `DrainResult` gains a `staged` field: items the judgment tier resolved as
+  "accept" that were not yet promoted in queue mode. These are reported as
+  resolved, not deferred.
+- `contentBodyLineCount` now uses the canonical `parseFrontmatter` for consistent
+  CRLF/BOM handling.
+
+### Workflow `when_to_use` Now Optional
+
+Workflow frontmatter no longer requires `when_to_use`. The field is still
+supported and recommended but the validator no longer rejects workflows that
+omit it.
 
 Release validation status for the final 0.8.0 cut: