adr: L2 harness runtime — self-built thin run-loop, no agent framework#33
Open
cyber-ayi wants to merge 1 commit into
Open
adr: L2 harness runtime — self-built thin run-loop, no agent framework#33cyber-ayi wants to merge 1 commit into
cyber-ayi wants to merge 1 commit into
Conversation
Resolve the ROADMAP `harness runtime` decision-gate (operator-acked 2026-06-25, multi-agent third-party harness-selection review). - decisions/adr-pending-a36d1d8e-l2-harness-runtime.md — self-built thin TS run-loop over the shipped Bridge; full rejected-alternatives audit covering ElizaOS / Mastra / AgentKit / Vercel AI SDK / VoltAgent / OpenAI Agents JS / LangGraph.js / OpenClaw / hermes-agent / AgentScope+LangGraph-py. Reuse at the library level (memory), never a runtime that owns the cognition loop. - exploration/prior-art-l2-l3.md — L2 runtime portion marked resolved → this ADR (L3 cognition survey stays input to T4). - ROADMAP.md — L2 row + `harness runtime` gate updated; supersedes the old ElizaOS/OpenClaw/Hermes lean. Session-Id: 019f02bb-6523-7115-a7b6-3fe0adfdda4c Agent: cc-rc-bot Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Resolves the ROADMAP
harness runtimedecision-gate. Operator-acked 2026-06-25 across a multi-agent third-party harness-selection review (vendor sources discounted; synthetic-pollution cross-verified).Decision
Self-built thin TypeScript run-loop over the shipped
harness/bridge/contract. No agent framework adopted. Reuse at the library level where it pays (memory: Letta/Mem0), never a runtime that owns the cognition loop.Common death-knell across every candidate: L3 brain is external Python over the JSON-Schema seam ⇒ a framework's LLM-orchestration value is bypassed (dead weight); general harnesses additionally carry cognition that invades the self-built L3.
Rejected (third-party audited)
ee/+ 2026-06 supply-chain poisoning; durable-workflow survivor mismatches a single-chain loopinngestcore below it@langchain/coreFiles
decisions/adr-pending-a36d1d8e-l2-harness-runtime.md(numbered on merge by adr-assign)exploration/prior-art-l2-l3.md— L2 portion resolved → this ADR (L3 survey stays T4 input)ROADMAP.md— L2 row +harness runtimegate updatedGateway/SDK decision (Matrix + mautrix-go) lands in a separate PR.
🤖 Generated with Claude Code