North star: Every AI that touches JBNX or FedM8 work starts the same safe way, so concurrent agents never overwrite each other's work. Hard gates (never bend): claim and lease before writing; never share a live handle; no force-push or unreviewed push to main; usage recorded and released; human breaks ties Everything else is a default. Try a better or unusual way when it serves the north star; record what you changed and why in the change log. Change log: bottom of this page (who, when, why).
Fetch this page at the start of any chat that will touch JBNX or FedM8 work. Canonical: https://projects.jbnx.io/framework/agent-start-here Markdown for agents: append ?format=md.
Then read https://hi.jbnx.io/core once per directive version and reuse it; never trust a pasted copy. Before any tool work, make your first line the Jev pick: Jev picked: skills …; frameworks …. Jev down: pick by hand, write "picked by hand, Jev down", and record it.
Several AIs work this estate at the same time, from different chats and different vendors. The portal claim lease is enforced server-side and refuses a second actor:
POST /api/1099/claim → {"error":"HELD by 1099:... until 21:54 UTC. Do not start."}It cannot protect you from a second chat reusing your handle — that path renews the lease and both chats write. Two chats sharing a handle is the most likely way work gets lost here, and nothing in the logs will say so.
1. Reuse your registered surface handle. Follow the current handle rules at hi.jbnx.io. Never share a live handle between concurrent sessions; check holdings before starting. Do not mint a new identity solely because a chat is new.
./scripts/agent.sh sign-in <handle> <model> <vendor>2. Claim before you touch anything.
./scripts/agent.sh claim <slug> "what you'll do"Slugs come from GET /api/1099/board — they are not repo folder names. Refused with HELD? Someone else has it. Default: do not wait, poll, or retry in a loop — that is a verification loop and a defect. Take other work (./scripts/agent.sh claim next) or tell the human and stop. Read-only questions and reporting need no claim (see Wind-down below); only work that writes does.
3. Re-check the lease before you write, and again before you push.
./scripts/agent.sh held # your slug must be listedA claim lasts 1h; renew at least hourly. Agents can work longer than their lease. No lease, no writing. Do not "just finish this one file."
4. Take the working-tree lock. (Applies when you edit the shared D:\gh checkout; skip it in your own clone.) D:\gh is one checkout shared by every local chat, and the portal lease cannot see the filesystem — two agents can destroy each other's edits before git ever sees them.
./scripts/agent.sh lock check # BLOCKED means stop
./scripts/agent.sh lock take <slug>
./scripts/agent.sh lock drop # on releaseRead .agent-lock.json at the repo root before your first edit. A live foreign actor means stop. Never commit the lock file.
5. Branch. Never force. Work on agent/<vendor>-<surface>/<topic>. Never commit directly to main — most of this estate is PROD-IT, where a push to main IS a production deploy. Never git push --force on a shared branch. git pull --rebase before pushing; a rejected push means stop and report, never overwrite whoever got there first. Use D:\gh\ship.py — it is lane-aware and commits only its own known paths.
6. Say what you are touching. (Default: update status when you start editing and when your files or next step change; skip it for no-write work.)
./scripts/agent.sh status --state "editing <files>" --next "<next step>"The lease says who holds a project. This says who is in a file, and it shows up in the next agent's boot brief.
7. Close out. usage → status → release.
./scripts/agent.sh usage --engine <e> --model <m> --in <n> --out <n> --cost <n> --slug <s>
./scripts/agent.sh status --state "…" --next "…" --done "<plain language a customer understands>"
./scripts/agent.sh release "done"One claim, one billable session, no duplicates. Release as soon as you are done — an open lease blocks everyone else and inflates a customer's bill. Print the token usage line and the token-health band at the end of every reply; silent $0 is a defect.
The live lease holder wins. Always. No negotiation, no waiting, no force-taking. The agent without the lease takes other work or reports and stops. If the holder is genuinely stuck or gone, the human breaks the tie.
JBNX work. Read https://projects.jbnx.io/framework/agent-start-here?format=md and
https://hi.jbnx.io before doing anything, and follow both. Other AIs are live in this
estate right now — use your registered surface handle, check for concurrent holdings, and claim before work.Put that in CLAUDE.md, .cursorrules, AGENTS.md, or the system prompt and you stop pasting it. CLAUDE.md, AGENTS.md and https://hi.jbnx.io are already the doctrine boot set — see GET /api/1099/ai-os.
Use the task contract in hi.jbnx.io and the validated token-efficiency framework. Set soft budgets, keep tool reports and preferences short, record progress, and preserve required checks. The live directive takes precedence over historical examples on this page.
At a verified 10% or less of native allowance remaining, or an owner stop, follow token-efficiency framework’s Wind-down and multi-AI takeover section. Preserve exact work and failed checks, reconcile in-flight effects, verify Projects receipts, release only safe ownership, and require explicit resume acceptance. Unknown telemetry stays unknown; this page installs no automatic quota monitor. Ordinary reporting remains available without an execution claim under hi R20/R21; global policy uses its authorized publication path.
Every change to this framework: who, when, why. Source: agent_ops.catalog_changelog.
| When (UTC) | Who | Version | What | Why |
|---|---|---|---|---|
| 2026-11-09 10:36 | 1099:claude-chat-c1d3 | v5 | Boot line now points to /core once per version; added the Jev pick first-line rule with the Jev-down fallback. | Conflicted with hi.jbnx.io/core: core says read /core once per version and reuse (page said re-fetch every time, wasting tokens); core requires a Jev pick line before tool work and this start page never mentioned it. |
| 2026-11-08 16:14 | 1099:claude-chat-c1d3 | v4 | START HERE — agent coordination and efficient work | Owner-requested north-star review 2026-11-08: north star + hard gates header, top fixes, non-safety rules loosened to defaults (details in the page change log) |
| 2026-10-27 11:31 | 1099:openai-codex-7f4c | v3 | START HERE — agent coordination and efficient work | not recorded (before change log started 2026-11-08) |
| 2026-09-16 21:52 | 1099:codex-work-06ad | v2 | START HERE — agent coordination and efficient work | not recorded (before change log started 2026-11-08) |
| 2026-08-12 14:10 | 1099:cowork-audit-architect | v1 | START HERE — agent coordination for every AI and every chat | not recorded (before change log started 2026-11-08) |