openclaw - 💡(How to fix) Fix RFC: `@openclaw/plugin-fleet-bridge` — typed dissent-preserving prior contract for multi-agent openclaw [1 comments, 2 participants]

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
GitHub stats
openclaw/openclaw#75788Fetched 2026-05-02 05:30:10
View on GitHub
Comments
1
Participants
2
Timeline
2
Reactions
2
Timeline (top)
closed ×1commented ×1
RAW_BUFFERClick to expand / collapse

This is the substrate for skill-panel-ask (Issue 1) and arguably the most upstream-worthy piece of work on this fork. It encodes a typed, append-only, dissent-preserving contract on top of the existing memory-graph plugin's SQLite store at ~/.openclaw/memory/graph.sqlite.

What it adds

Two new node types (additive to memory-graph):

  • prior-capsule: hash-chained shared world-state. parent_hash chains capsules; claim_refs index folded claims; dissent_refs index preserved dissents; evidence_index ties claims to evidence node ids.
  • posterior-delta: typed reply from a voice keyed against a specific prior. Includes claims (with confidence + evidence refs), dissent (against specific claims), proposed mutations, family + surface, and a self-doubt note.

Plus compaction-record (three-signature ceremony: proposer/approver/signer) and dissent-record (preserved across compactions as schema invariant).

Schema-level invariants (enforced in code, not policy)

  1. Append-only — prior-capsule, posterior-delta, compaction-record, dissent-record are all immutable after insert. Plugin rejects any UPDATE or DELETE on these node types.

  2. Compaction cannot drop dissent — a CompactionRecord that omits any dissent id from its constituent deltas is rejected at validation.

  3. Three-signature compaction — proposer + approver + signer must all be distinct voice ids. (In my Cabinet design: Compiler proposes, Court approves, Notary signs — three separate openclaw agents.)

  4. Read markers per voice — each prior tracks read_markers[voice_id] so a voice can see "the prior advanced since I last read; here are my new deltas to apply."

  5. Cross-family discipline at synthesis time — same rule from Issue 1, but applied to compaction-record signing too: the approver's family ≠ majority contributor family.

Why now

Other multi-agent frameworks don't have a typed dissent-preserving prior contract. They stuff strings into a vector DB and call it memory. The prior+delta contract is what makes openclaw a governed agentic OS rather than another LLM router.

Full spec at ~/.openclaw/workspace/morning-summary/prior-delta-schema.md (this fork). 32-40 hour estimate to ship the plugin including invariant validators, three-signature ceremony, dissent preservation suite, and migration path from existing markdown prior capsules.

Happy to send a PR. Would appreciate:

  • Confirmation the memory-graph plugin is the right home (vs a standalone plugin)
  • Guidance on schema migration tooling (does memory-graph have a versioned migration runner I should hook into?)

extent analysis

TL;DR

The developer should integrate the new node types and schema invariants into the existing memory-graph plugin, ensuring compatibility and proper migration.

Guidance

  • Review the memory-graph plugin's architecture to confirm it's the suitable home for the new features, considering factors like data structure and scalability.
  • Investigate the plugin's versioned migration runner, if available, to determine the best approach for schema migration and integration of the new node types.
  • Ensure the implementation of schema invariants, such as append-only and compaction rules, is correctly enforced in the code to maintain data integrity.
  • Plan for a thorough testing process to verify the correct functioning of the new features and their interaction with the existing plugin.

Notes

The provided information lacks specific technical details about the plugin's current implementation and migration tools, so a more detailed guidance requires additional context or clarification.

Recommendation

Apply workaround: Integrate the new features into the memory-graph plugin while ensuring compatibility and proper migration, as this approach allows for a more streamlined and maintainable solution.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING

openclaw - 💡(How to fix) Fix RFC: `@openclaw/plugin-fleet-bridge` — typed dissent-preserving prior contract for multi-agent openclaw [1 comments, 2 participants]