New @logicsrc/plugin-credential-sharing: a provider-neutral secret-sync engine
with env/.env, Doppler, Railway, and GitHub Secrets adapters behind one
CredentialProvider contract.
- engine: inspect -> diff -> plan -> approve -> sync -> rollback -> audit/export
- dry-run is the default for sync; --approve writes; destructive changes gated
- fingerprint-based diffs (salted SHA-256); raw values never printed or stored in
plans/runs/audit; rollback pre-image kept in a 0600 .logicsrc vault (gitignored)
- github-secrets is write-only for values (sealed-box via libsodium), so it cannot
be a sync source or value-restoring rollback target
- CLI: real `logicsrc credentials <providers|inspect|diff|plan|approve|sync|
rollback|audit|export>` (replaces the prior stub)
- 4 JSON schemas registered in @logicsrc/validators
- flip logicsrc.com/credential-sharing band from coming-soon to available
- 37 tests pass; full env->env lifecycle verified; artifacts schema-validate
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add files/publishConfig(public)/repository/keywords + prepublishOnly build;
exclude *.test.ts from the tsc build so dist (and the published tarball) ships
only library code. Package size 17 kB, 38/38 tests still green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
withX402() wraps the Web handler so unpaid calls get a standard HTTP 402
challenge (x402Version + accepts[]) and paid calls fall through to the agent.
Payment verification is injectable (host wires CoinPay/x402); CORS preflight
passes through untouched; custom challenge responses supported. README documents
the optional c0mpute.com GPU backend.
4 new tests; 38/38 pass.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
createC0mputeModel() turns a connected c0mpute.com account (apiKey + OpenAI-
compatible base URL) into a LangChain chat model, usable as the model for
createDeepAgentRunner or the chatModel for the judge/router — so agent inference
can run on community-shared GPUs. Opt-in; default stays the hosted provider.
agentswarm performs no OAuth: the host's c0mpute connector supplies credentials
(resolveC0mputeConnector / c0mputeConnectorFromEnv normalize them). DeepAgent
runner now accepts a model instance, not just a provider string.
7 new tests (defaults, env mapping, not-connected + missing-peer errors); 34/34 pass.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
BudgetLedger interface + InMemoryBudgetLedger reference impl (reserve/settle/
release with refund of unused holds; mirrors the b1dz budget+ledger concept).
withBudget() wraps any SwarmRunner: reserves up front (throws SwarmError 402 when
out of funds, before running), settles actual cost on success, releases on error.
AgentIdentity DID aligns with @logicsrc/agentstack.
8 new tests; 27/27 pass.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
createSwarm() composes named agents into one SwarmRunner: a router picks the
first agent; any agent hands control to a named peer via `HANDOFF: <name>`,
bounded by maxHandoffs. createLLMRouter() is a cheap-model router. Composes
with the rubric loop and HTTP handler since a swarm is itself a SwarmRunner.
6 new tests (routing, handoff, maxHandoffs cap, unknown-peer/unknown-route
fallbacks, empty-agents guard); 19/19 pass.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
@logicsrc/agentswarm: a self-hosted, framework-agnostic multi-agent runtime
that each app mounts on its own route (e.g. tronbrowser.dev/swarm).
M1 — core:
- createSwarmHandler(): Web (Request)=>Response handler, CORS + validation
- SwarmRunner: injectable engine interface
- createDeepAgentRunner(): deepagents (createDeepAgent) adapter; deepagents +
@langchain/langgraph are optional peers loaded via dynamic import so the core
builds/tests with zero heavy deps
- SwarmError + onRequest gate: seam for x402 metering / auth (402/403)
M3 — rubric self-check (deepagents RubricMiddleware is Python-only, so ported
at the runner layer):
- createRubricRunner(): grades output via an injectable judge and revises until
it passes or maxIterations; passes through untouched when no rubric
- createLLMJudge(): cheap-model grader (lazy langchain initChatModel)
13/13 vitest pass; tsc clean; examples/server.mjs demo verified.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
M1 of the ANS SDK (docs/ans-sdk.md):
- name: parse/format ans://v<semver>.<agent>.<domain>
- cbor: minimal RFC 8949 codec for the COSE_Sign1 subset
- verify/merkle: RFC 6962 leaf/node hashing, tree build, inclusion-proof
generation + verification
- verify/es256 + cose: COSE_Sign1 build/parse, Sig_structure, ES256 (WebCrypto)
- verify/rootkeys: kid -> verifier key (JWKS entries; sumdb-note is M2)
- verify: verifyReceipt() + verifyResolution() (signature + inclusion proof +
name binding), pure and offline
- client: AnsClient resolve/rootKeys/register/status over injectable fetch,
with a DnsApplier hook for verify-dns
Tests (24) cover name parsing, CBOR round-trips/vectors, RFC 6962 proofs, a full
ES256+Merkle receipt round-trip with positive/negative cases, and the client
against mocked fetch. Wired into the root build script.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Design doc for the LogicSRC TS SDK for the Agent Name Service (ANS): resolver +
offline transparency-log verifier + registration client + coinpay-DID bridge.
Covers scope, package layout, full API surface, the ans-verify algorithm,
milestones (M1 verifier -> M2 register -> M3 DID bridge), and risks.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Product requirements for @logicsrc/agentstack: summary, goals/non-goals,
users, current v0.1.0 state, functional/non-functional requirements,
roadmap (M1–M6), success metrics, and open questions. Grounded in the
existing in-memory coordinator and the agentstack OpenSpec capability.
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
The agentmail workspace was added to package.json but the lockfile was not
regenerated, so CI's `npm ci` failed (EUSAGE: missing from lock file). Regenerate.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
AgentGit is a thin, DID-gated source-collaboration layer over a backend
forge (default Forgejo at git.profullstack.com, BBS-members-only) — not a
new git host. M1 implements the contract and engines:
- forge/adapter.ts: ForgeAdapter interface (only forge-specific surface)
- forge/forgejo.ts: ForgejoAdapter over Forgejo/Gitea REST v1 (injectable
fetch, typed errors), incl. ensureUser for member provisioning
- access.ts: gateAccess DID membership gate (owner/role/visibility)
- merge-policy.ts: evaluateMergePolicy pure engine (reviews, reputation
floor, checks, escrow, merge method, agent-merge toggle)
- service.ts: AgentGitService ties gate + policy to the adapter; refuses
policy-failing merges; provisionMember hook for AgentBBS
- schemas: logicsrc-repo + logicsrc-pull-request, registered in
@logicsrc/validators with fixtures
- docs/agentgit.md spec; plugin wired into root build (default/disabled)
27 vitest tests pass; full monorepo build green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
'value in schemas' walks the prototype chain, so inherited keys like
"toString" passed assertSchemaKind and then crashed ajv.compile with a
misleading 'schema must be object or boolean' error (e.g. via the CLI).
Use Object.hasOwn for an own-property check.
Adds unit tests rejecting prototype keys and asserting every real
schema kind still passes.
Fixes#13
