akm-cli 0.9.29 → 0.9.31
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/CHANGELOG.md +41 -0
- package/README.md +10 -10
- package/ROADMAP.md +112 -0
- package/dist/commands/improve/improve-cli.js +33 -12
- package/dist/commands/sources/plugin-upgrade.js +46 -23
- package/dist/core/file-lock.js +10 -1
- package/dist/llm/client.js +33 -3
- package/dist/scripts/akm-migrate-node.js +5 -1
- package/dist/scripts/akm-migrate.js +5 -1
- package/docs/README.md +42 -41
- package/docs/agents/README.md +15 -0
- package/docs/agents/agent-install.md +14 -0
- package/docs/architecture/README.md +67 -0
- package/docs/architecture/adapters.md +272 -0
- package/docs/architecture/akm-architecture-decision-history.md +2055 -0
- package/docs/architecture/akm-core-principles.md +126 -0
- package/docs/architecture/architecture.md +579 -0
- package/docs/architecture/comparisons/claude-code-vs-akm-workflows-full.md +550 -0
- package/docs/architecture/decisions/0001-no-interpolation-attached-structured-context.md +72 -0
- package/docs/architecture/decisions/0002-unit-reuse-and-input-hash-scope.md +142 -0
- package/docs/architecture/decisions/0003-child-env-allowlist-and-provenance.md +189 -0
- package/docs/architecture/decisions/0004-task-input-contract-and-flag-coercion.md +97 -0
- package/docs/architecture/decisions/0005-task-result-vocabulary-and-legacy-read-mapping.md +108 -0
- package/docs/architecture/decisions/0006-task-source-version-routing.md +124 -0
- package/docs/architecture/decisions/0007-workflow-composition-bounds.md +83 -0
- package/docs/architecture/decisions/0008-task-binding-normalization.md +88 -0
- package/docs/architecture/decisions/0009-child-run-publication-column-parity.md +80 -0
- package/docs/architecture/decisions/0010-driver-protocol-cut.md +36 -0
- package/docs/architecture/decisions/0011-engine-run-loop-invariants.md +125 -0
- package/docs/architecture/decisions/README.md +62 -0
- package/docs/architecture/improvement.md +440 -0
- package/docs/architecture/internals/classification.md +154 -0
- package/docs/architecture/internals/fresh-host-rebuild-runbook.md +40 -0
- package/docs/architecture/internals/functional-contract-patterns.md +340 -0
- package/docs/architecture/internals/health-advisories.md +110 -0
- package/docs/architecture/internals/improve-workflow.md +713 -0
- package/docs/architecture/internals/indexing.md +498 -0
- package/docs/architecture/internals/registry-network-boundary.md +57 -0
- package/docs/architecture/internals/search.md +249 -0
- package/docs/architecture/internals/storage-locations.md +943 -0
- package/docs/architecture/persisted-data-compat.md +59 -0
- package/docs/architecture/reviews/env-secret-access.md +319 -0
- package/docs/architecture/runtime-boundary-design.md +153 -0
- package/docs/architecture/specs/0.9.0-close-out-plan.md +361 -0
- package/docs/architecture/specs/0.9.0-decisions.md +528 -0
- package/docs/architecture/specs/0.9.0-docs-code-drift-register.md +745 -0
- package/docs/architecture/specs/0.9.0-open-items-register.md +135 -0
- package/docs/architecture/specs/0.9.0-public-api-issue-backlog.md +740 -0
- package/docs/architecture/specs/0.9.0-release-surface-review.md +636 -0
- package/docs/architecture/specs/agent-command-engine-model-design.md +513 -0
- package/docs/architecture/specs/akm-0.9.0-bundle-adapter-spec.md +501 -0
- package/docs/architecture/specs/akm-0.9.0-ref-grammar-decision.md +291 -0
- package/docs/architecture/specs/akm-format-neutral-bundle-workspace-spec.md +2059 -0
- package/docs/architecture/specs/di-seams-plan.md +539 -0
- package/docs/architecture/specs/driver-protocol-keep-or-cut.md +275 -0
- package/docs/architecture/specs/fact-asset-type.md +142 -0
- package/docs/architecture/specs/improve-collapse-churn-detector-design.md +634 -0
- package/docs/architecture/specs/okf-support.md +148 -0
- package/docs/architecture/specs/ref.md +329 -0
- package/docs/architecture/specs/reviews/task-workflow-unification/README.md +72 -0
- package/docs/architecture/specs/reviews/task-workflow-unification/findings-consolidated.md +152 -0
- package/docs/architecture/specs/reviews/task-workflow-unification/library-eval-alternatives.md +352 -0
- package/docs/architecture/specs/reviews/task-workflow-unification/library-eval-bunqueue.md +300 -0
- package/docs/architecture/specs/reviews/task-workflow-unification/library-eval-openworkflow.md +402 -0
- package/docs/architecture/specs/reviews/task-workflow-unification/opus-review-1.md +758 -0
- package/docs/architecture/specs/reviews/task-workflow-unification/opus-review-2.md +645 -0
- package/docs/architecture/specs/reviews/task-workflow-unification/opus-review-3.md +602 -0
- package/docs/architecture/specs/reviews/task-workflow-unification/panel-r1-adversarial.md +27 -0
- package/docs/architecture/specs/reviews/task-workflow-unification/panel-r1-architecture.md +27 -0
- package/docs/architecture/specs/reviews/task-workflow-unification/panel-r1-coverage.md +74 -0
- package/docs/architecture/specs/reviews/task-workflow-unification/panel-r1-security.md +15 -0
- package/docs/architecture/specs/reviews/task-workflow-unification/panel-r2-adversarial.md +132 -0
- package/docs/architecture/specs/reviews/task-workflow-unification/panel-r2-verification.md +147 -0
- package/docs/architecture/specs/stash-conventions-code-spec.md +275 -0
- package/docs/architecture/specs/stash-organization-conventions.md +377 -0
- package/docs/architecture/specs/task-workflow-format-unification.md +721 -0
- package/docs/architecture/specs/workflow-engine-buy-vs-build.md +231 -0
- package/docs/architecture/specs/workflow-format-unification.md +437 -0
- package/docs/architecture/testing/manual-testing-checklist.md +3420 -0
- package/docs/architecture/testing/okf-v0.2-conformance-runbook.md +924 -0
- package/docs/architecture/testing/task-workflow-manual-tests.md +1817 -0
- package/docs/architecture/testing/testing-workflow.md +446 -0
- package/docs/architecture/workflow-engine.md +274 -0
- package/docs/guides/README.md +33 -0
- package/docs/guides/author-bundles.md +310 -0
- package/docs/guides/author-workflows.md +520 -0
- package/docs/guides/bundles.md +321 -0
- package/docs/guides/capture-knowledge.md +136 -0
- package/docs/guides/claude-code-vs-akm-workflows.md +30 -0
- package/docs/guides/concepts.md +93 -0
- package/docs/guides/discover-and-load.md +158 -0
- package/docs/guides/getting-started.md +240 -0
- package/docs/guides/improve-the-library.md +246 -0
- package/docs/guides/knowledge-management.md +18 -0
- package/docs/guides/recipes/headless-install.md +97 -0
- package/docs/guides/recipes/website-source.md +28 -0
- package/docs/guides/run-workflows.md +279 -0
- package/docs/guides/scheduling.md +123 -0
- package/docs/guides/use-with-any-agent.md +129 -0
- package/docs/guides/wikis.md +81 -0
- package/docs/maintainers/README.md +19 -0
- package/docs/maintainers/curate-workmap.md +336 -0
- package/docs/maintainers/eval.md +3 -0
- package/docs/maintainers/local-development.md +108 -0
- package/docs/maintainers/release-checklist.md +103 -0
- package/docs/migration/README.md +2 -2
- package/docs/migration/release-notes/0.0.13.md +1 -1
- package/docs/migration/release-notes/0.1.0.md +1 -1
- package/docs/migration/release-notes/0.2.0.md +1 -1
- package/docs/migration/release-notes/0.3.0.md +1 -1
- package/docs/migration/release-notes/0.5.0.md +1 -1
- package/docs/migration/release-notes/0.6.0.md +1 -1
- package/docs/migration/release-notes/0.7.0.md +1 -1
- package/docs/migration/release-notes/0.7.3.md +2 -2
- package/docs/migration/release-notes/0.7.4.md +1 -1
- package/docs/migration/release-notes/0.7.5.md +1 -1
- package/docs/migration/release-notes/0.8.0.md +1 -1
- package/docs/migration/release-notes/0.9.0.md +1 -1
- package/docs/migration/release-notes/0.9.14.md +1 -1
- package/docs/migration/release-notes/0.9.15.md +1 -1
- package/docs/migration/release-notes/0.9.16.md +1 -1
- package/docs/migration/release-notes/0.9.17.md +1 -1
- package/docs/migration/release-notes/0.9.19.md +1 -1
- package/docs/migration/release-notes/0.9.2.md +1 -1
- package/docs/migration/release-notes/README.md +27 -31
- package/docs/migration/v0.5-to-v0.6.md +759 -0
- package/docs/migration/v0.8-to-v0.9.md +1 -1
- package/docs/plans/0.9.2-agent-command-workflow-plan.md +423 -0
- package/docs/posts/README.md +30 -0
- package/docs/posts/akm-command-tour.md +495 -0
- package/docs/posts/akm-logo-sized.webp +0 -0
- package/docs/posts/belief-aware-memory-12.md +139 -0
- package/docs/posts/config-v2-lessons-16.md +191 -0
- package/docs/posts/improve-loop-autonomous-10.md +225 -0
- package/docs/posts/improve-pipeline-debugging-0803.md +179 -0
- package/docs/posts/intro-part-01.md +76 -0
- package/docs/posts/intro-part-02.md +221 -0
- package/docs/posts/intro-part-03.md +190 -0
- package/docs/posts/intro-part-04.md +221 -0
- package/docs/posts/llm-mode-reflect-13.md +166 -0
- package/docs/posts/openpalm-built-on-akm.md +86 -0
- package/docs/posts/progressive-disclosure-06.md +129 -0
- package/docs/posts/proposal-safety-net-15.md +112 -0
- package/docs/posts/release-0.5.0.md +247 -0
- package/docs/posts/release-0.6.0.md +117 -0
- package/docs/posts/release-0.7.0.md +159 -0
- package/docs/posts/release-0.8.0.md +332 -0
- package/docs/posts/scattered-skills-05.md +100 -0
- package/docs/posts/task-assets-persistent-workflows-11.md +222 -0
- package/docs/posts/team-skills-07.md +198 -0
- package/docs/posts/wiki-agent-knowledge-08.md +223 -0
- package/docs/posts/workflows-vaults-09.md +321 -0
- package/docs/product-surface.md +54 -0
- package/docs/reference/README.md +1 -2
- package/docs/reference/asset-types.md +168 -0
- package/docs/reference/cli.md +5 -6
- package/docs/reference/env-and-secrets.md +126 -0
- package/docs/reference/memory.md +124 -0
- package/docs/reference/refs.md +156 -0
- package/docs/reference/registry.md +489 -0
- package/docs/reference/website-sources.md +141 -0
- package/package.json +12 -15
|
@@ -0,0 +1,126 @@
|
|
|
1
|
+
# akm Core Principles
|
|
2
|
+
|
|
3
|
+
akm is a portable capability library for AI agents — give every coding agent
|
|
4
|
+
the capabilities your team has already built. It indexes existing agent
|
|
5
|
+
assets in place, loads only what a task needs, packages capabilities into
|
|
6
|
+
shareable bundles, improves the library through reviewable proposals, and
|
|
7
|
+
runs durable workflows, locally and without tying the library to one
|
|
8
|
+
assistant.
|
|
9
|
+
|
|
10
|
+
This document lists the design constraints that keep that library
|
|
11
|
+
predictable.
|
|
12
|
+
|
|
13
|
+
## What akm Does
|
|
14
|
+
|
|
15
|
+
An agent has a task. Across configured sources (local filesystem paths and
|
|
16
|
+
cache-backed git/website/npm mirrors) and registry catalogs, akm helps it
|
|
17
|
+
discover assets such as scripts, skills, commands, agents, knowledge docs,
|
|
18
|
+
workflows, env files, secrets, and wiki pages.
|
|
19
|
+
|
|
20
|
+
Core flow:
|
|
21
|
+
|
|
22
|
+
```text
|
|
23
|
+
connect -> index -> curate -> show -> use/run -> feedback -> proposal
|
|
24
|
+
```
|
|
25
|
+
|
|
26
|
+
Not every task uses every stage.
|
|
27
|
+
|
|
28
|
+
## Principles
|
|
29
|
+
|
|
30
|
+
### 1. Capabilities remain portable files
|
|
31
|
+
|
|
32
|
+
Capabilities are ordinary files on disk — skills, scripts, workflows, agent
|
|
33
|
+
definitions, instructions, memories, knowledge docs. akm indexes what's
|
|
34
|
+
already there; the source files remain the source of truth, and akm's index
|
|
35
|
+
is a derived, rebuildable view rather than a copy of record.
|
|
36
|
+
|
|
37
|
+
### 2. Materialize sources locally and use one retrieval layer
|
|
38
|
+
|
|
39
|
+
Local filesystem paths and cache-backed git/website/npm mirrors all
|
|
40
|
+
materialize into one local index, so `search`, `curate`, and `show` work the
|
|
41
|
+
same way regardless of where a capability originated. Registries stay a
|
|
42
|
+
conceptually separate, read-only catalog of installable bundles: registry
|
|
43
|
+
results live in `registryHits`, never merged into source `hits`.
|
|
44
|
+
|
|
45
|
+
### 3. Every token must earn its place
|
|
46
|
+
|
|
47
|
+
Default output should stay lean. Search and curate are for choosing; show is
|
|
48
|
+
for using.
|
|
49
|
+
|
|
50
|
+
- **Search is a menu.** Default search output should expose only enough to
|
|
51
|
+
choose the next asset. In the current CLI that usually means:
|
|
52
|
+
- `brief`: `type`, `name`, `ref`, `action`, `estimatedTokens`, `keys` —
|
|
53
|
+
`ref` is deliberately present at the leanest level so an agent can run
|
|
54
|
+
`akm show <ref>` without asking for fuller detail
|
|
55
|
+
- `normal`: adds `description` and `score`
|
|
56
|
+
|
|
57
|
+
Richer provenance/debug fields belong behind fuller detail modes.
|
|
58
|
+
|
|
59
|
+
- **Show is a dispatch envelope.** Show should return the payload that lets
|
|
60
|
+
the consumer act:
|
|
61
|
+
- script execution hints
|
|
62
|
+
- skill instructions
|
|
63
|
+
- command templates
|
|
64
|
+
- agent prompts
|
|
65
|
+
- knowledge/wiki content
|
|
66
|
+
- workflow steps and parameters
|
|
67
|
+
- env key names without secret values
|
|
68
|
+
- secret metadata without the secret value
|
|
69
|
+
|
|
70
|
+
### 4. Discovery chooses; show delivers
|
|
71
|
+
|
|
72
|
+
Search and curate should not accumulate show-level detail. `full` detail
|
|
73
|
+
modes can expose more metadata, but the base mental model stays:
|
|
74
|
+
|
|
75
|
+
```text
|
|
76
|
+
search/curate decide
|
|
77
|
+
show delivers
|
|
78
|
+
filesystem is optional depth
|
|
79
|
+
```
|
|
80
|
+
|
|
81
|
+
### 5. Refs are opaque durable handles
|
|
82
|
+
|
|
83
|
+
Consumers should treat refs as opaque lookup handles. The current wire
|
|
84
|
+
format is `[bundle//]conceptId[#fragment]`, but agents should pass refs
|
|
85
|
+
through rather than parse them. See `docs/architecture/specs/ref.md`.
|
|
86
|
+
|
|
87
|
+
### 6. Execution requires an explicit supported surface
|
|
88
|
+
|
|
89
|
+
akm retrieves every supported capability type. It directly orchestrates
|
|
90
|
+
defined execution surfaces such as workflows, agent dispatch, tasks, and
|
|
91
|
+
guarded subprocess injection. It does not blindly execute arbitrary indexed
|
|
92
|
+
content merely because that content appears in search results.
|
|
93
|
+
|
|
94
|
+
### 7. Improvement is evidence-driven and reviewable
|
|
95
|
+
|
|
96
|
+
Feedback and usage signal drive `akm improve`, which produces reviewable,
|
|
97
|
+
diffable proposals rather than silent rewrites. Accept, reject, and revert
|
|
98
|
+
stay human- or policy-gated, and proposals target only writable bundles.
|
|
99
|
+
|
|
100
|
+
### 8. Writes are destination-aware and fail closed
|
|
101
|
+
|
|
102
|
+
Writes target only writable bundles/adapters. Read-only adapters and
|
|
103
|
+
protected env/secret values are never silently written through — a write
|
|
104
|
+
aimed at a non-writable destination fails rather than falling back to an
|
|
105
|
+
unintended target.
|
|
106
|
+
|
|
107
|
+
### 9. Output serves agents first
|
|
108
|
+
|
|
109
|
+
The default consumer is structured-output automation. JSON-first and
|
|
110
|
+
concise detail levels are the right defaults.
|
|
111
|
+
|
|
112
|
+
### 10. Complexity belongs behind indexing and source management
|
|
113
|
+
|
|
114
|
+
The hard parts should stay inside indexing, source resolution, registry
|
|
115
|
+
install, and provider plumbing, not in the hot path from `search`/`curate`
|
|
116
|
+
to `show`/`use`.
|
|
117
|
+
|
|
118
|
+
## What akm Does Not Do
|
|
119
|
+
|
|
120
|
+
- blindly execute arbitrary indexed content just because it appears in
|
|
121
|
+
search results — akm orchestrates only defined execution surfaces
|
|
122
|
+
(workflows, agent dispatch, tasks, guarded subprocess injection)
|
|
123
|
+
- expose secret/env values through `show`
|
|
124
|
+
- replace live-service integrations such as MCP — akm complements MCP and
|
|
125
|
+
assistant-native skills rather than replacing them
|
|
126
|
+
- require agents to understand source layouts or provider internals
|