I always say that the difference between OpenClaw vs Hermes is simply project vs product. OpenClaw started as Peter’s project and get some crazy traction, many people onboard and keeps the hobby project’s “ship fast, break thinks” vibe. On the other hand, Hermes Agent feels more like a product that is backed by a company that knows how to actually deliver products. Releases are balanced, addressing the biggest pain points instead of everything everywhere all at once. The work with the community is also different, more mature, but also more marketingish.
agreed. they have definitely solidified their wedge. looking forward to seeing what they do next and if they can keep up with openclaw long term. essentially it’s a fork of openclaw that has been modified, you can import the Hermes features into an openclaw setup and it run the same.
I always say that the difference between OpenClaw vs Hermes is simply project vs product. OpenClaw started as Peter’s project and get some crazy traction, many people onboard and keeps the hobby project’s “ship fast, break thinks” vibe. On the other hand, Hermes Agent feels more like a product that is backed by a company that knows how to actually deliver products. Releases are balanced, addressing the biggest pain points instead of everything everywhere all at once. The work with the community is also different, more mature, but also more marketingish.
agreed. they have definitely solidified their wedge. looking forward to seeing what they do next and if they can keep up with openclaw long term. essentially it’s a fork of openclaw that has been modified, you can import the Hermes features into an openclaw setup and it run the same.
Here’s a project update I just got from Hermes. Cronjob Response: project-status-brief
(job_id: 0a1183af5e6f)
-------------
I now have all the data. Here's the compiled brief.
📊 PROJECT STATUS BRIEF — 2026-04-29 18:05 CT
━━━ S-001: Historical Data & Concept Lab ━━━
Phase 1 — Data Engine: 🟡 YELLOW
Equities: 26,052 rows | 2024-04-29 → 2026-04-28 (2yr, 29 tickers)
Crypto: 5,805 rows | 2024-04-29 → 2026-04-29 (9/10 symbols; MATIC no source)
ETFs: 3,006 rows | 2024-04-29 → 2026-04-28 (6 tickers, data present)
Events: All 3 DBs — 0 events rows ❌ (feature not populated)
Metadata: Stale — all equities/ETFs show "no_data" despite rows in DB
Timer: Triggered at 17:01, exit 0/SUCCESS. Next fire: 17:00 CT tomorrow.
Assessment: Core OHLC data is solid across all 3 asset classes.
Two gaps: (1) metadata.json wasn't updated to reflect actual data state
after the Hermes backfill fix; (2) events table never written to any DB.
Crypto missing MATIC (no source data). Overall data pipeline works but
verification/ENRICHMENT step is incomplete.
Phase 2 — Concept Lab: Inbox has Scock's lab plan handoff (16:56)
received. Not yet reviewed.
━━━ S-002: Janus/Sulu (Unified Recall) ━━━
Phase: DESIGN / SPEC'D — 🟡 YELLOW
Scock response status:
✅ RECEIPTs received for: chain-of-command, corrective-actions,
sulu-runtime-model, morning-diagnostics-v02
✅ New handoffs from Scock (post-receipt):
→ Compass recall design spec (15:44) — 172-line design doc
→ GMGN spec (16:30)
→ GMGN crypto integration (17:25)
→ S-001 Phase 2 lab plan (16:56)
Status: Design phase active. Scock is producing specs. Sulu's runtime
model is resolved (MCP toolset, not interceptor or cron). Compass
unified recall design is in progress. PROJECTS.md is still empty —
needs populating with active project entries.
Outgoing: 2 lingering DRAFT files never shipped (historical data lab
plan + schema). Scock needs to ship or kill these.
━━━ H-001: Sulu Agent On-Ramp ━━━
Phase: SPEC'D — 🟢 GATE CLOSED
Spec exists: projects-office/on-ramp-sulu.md (v0.1, 2,718 bytes)
Build status: NOT BUILT — no AGENTS.md, no systemd service, no MCP
registration. Sulu doesn't exist as a runtime entity yet.
Gate: Closed. Waiting on spec approval → Scock implementation design
→ Kirk build.
━━━ System Health ━━━
Dashboard (HTTP): Unable to connect on :8080/:8082 — both gateways
confirm active via systemd (non-HTTP protocol ports).
Hermes Gateway: ✅ active/running
OpenClaw Gateway: ✅ active/running
Timers: 16 wb-* timers — mostly healthy
❌ wb-crypto-digest-4h.service — FAILED (exit code 1) at 16:25:58 CT
Timer itself is still active/waiting. Next fire at 20:00 CT.
Needs investigation — likely script error in the 4h digest.
✅ wb-historical-data-ingest — last run success at 17:01 CT
✅ All other services: inactive/dead as expected (one-shot timers)
━━━ Summary ━━━
S-001: 🟡 Data is in the DB but verification metadata and events need work
S-002: 🟡 Active design phase, Scock producing specs, good momentum
H-001: 🟢 Spec exists, awaiting build trigger
Sys: 🔴 crypto-digest-4h failed — needs diagnosis before next fire
To stop or manage this job, send me a new message (e.g. "stop reminder project-status-brief").