layered-soul/memories/curated/agent-roster.md

# Agent Roster

| Agent  | Home Host | Harness      | Model    | Role                                        |
| ------ | --------- | ------------ | -------- | ------------------------------------------- |
| Hermes | debby     | Hermes Agent | DeepSeek | Orchestrator, primary dev agent             |
| Zot    | debby     | Zot RPC      | GLM-5.1  | Coding specialist, media workflows          |
| Claude | domedog   | Claude Code  | Claude   | Independent verification, secondary builder |
| Codex  | osa       | Codex CLI    | GPT      | FreeBSD 15 validation, ISO builds           |

## Multi-agent group

"My Debby" Telegram group: @hermes_samob_bot + @zot_samob_bot. Hermes orchestrates, delegates coding/media to Zot via structured delegation.

## Secrets

All agent API keys live in Vaultwarden at vault.smilepowered.org (BW_SERVER). Bootstrap credentials (BW_CLIENTID, BW_CLIENTSECRET, BW_PASSWORD) are the only secrets in `.env` files — everything else is fetched from the vault at runtime.

## Verification

Claude (domedog) independently verifies Hermes's claims — test counts, clippy status, crate counts. Do not overstate results. Use exact `cargo test --workspace` output.
Populate layered-soul: identity, memories, skills, plan (Hermes & Sam) - SOUL.md: full agent identity, operating principles, voice - IDENTITY.md: runtime identity, hosts, boundaries - USER.md: operator context imported from hermes-soul - AGENTS.md: actual operating rules, infrastructure, quick reference - memories/curated/: 5 topics (tailscale, forgejo, agents, projects, vaultwarden) - skills/: 9 cross-harness skills imported from hermes-soul after review - docs/PLAN-CONFIGURE-PRIVATE-REPO.md: configuration plan - Validate: passes clean 2026-06-14 00:21:26 +02:00			`# Agent Roster`

docs: apply Prettier to current markdown (Sam & Codex) Normalize markdown formatting after the latest main updates.\n\nChecks: python3 scripts/layered_soul.py validate .; npx --yes prettier@3 --check '*/.md'; git diff --check. 2026-06-14 01:48:32 +02:00			`\| Agent \| Home Host \| Harness \| Model \| Role \|`
			`\| ------ \| --------- \| ------------ \| -------- \| ------------------------------------------- \|`
			`\| Hermes \| debby \| Hermes Agent \| DeepSeek \| Orchestrator, primary dev agent \|`
			`\| Zot \| debby \| Zot RPC \| GLM-5.1 \| Coding specialist, media workflows \|`
			`\| Claude \| domedog \| Claude Code \| Claude \| Independent verification, secondary builder \|`
			`\| Codex \| osa \| Codex CLI \| GPT \| FreeBSD 15 validation, ISO builds \|`
Populate layered-soul: identity, memories, skills, plan (Hermes & Sam) - SOUL.md: full agent identity, operating principles, voice - IDENTITY.md: runtime identity, hosts, boundaries - USER.md: operator context imported from hermes-soul - AGENTS.md: actual operating rules, infrastructure, quick reference - memories/curated/: 5 topics (tailscale, forgejo, agents, projects, vaultwarden) - skills/: 9 cross-harness skills imported from hermes-soul after review - docs/PLAN-CONFIGURE-PRIVATE-REPO.md: configuration plan - Validate: passes clean 2026-06-14 00:21:26 +02:00
			`## Multi-agent group`

			`"My Debby" Telegram group: @hermes_samob_bot + @zot_samob_bot. Hermes orchestrates, delegates coding/media to Zot via structured delegation.`

			`## Secrets`

			All agent API keys live in Vaultwarden at vault.smilepowered.org (BW_SERVER). Bootstrap credentials (BW_CLIENTID, BW_CLIENTSECRET, BW_PASSWORD) are the only secrets in `.env` files — everything else is fetched from the vault at runtime.

			`## Verification`

			Claude (domedog) independently verifies Hermes's claims — test counts, clippy status, crate counts. Do not overstate results. Use exact `cargo test --workspace` output.