OpenClaw Loop Infrastructure Repair + Agent Autonomy Engine (2-Week Sprint)

We operate an agent fleet orchestration platform (OpenClaw) with 10+ products and 9 autonomous agents. Currently have 109 active automated workflows (loops).

YOUR MISSION (2-week intensive sprint):
1. Audit and repair all 109 loops against our spec
2. Create a unified test framework (bash-based, zero AI cost)
3. Verify all 109 loops healthy and tested (109/109 PASS required)
4. Design and implement an Autonomy Engine so agents discover findings AND act on them (not just report)
5. Build 4-6 new autonomy-focused loops (customer care, refund processing, self-improvement, lead nurture, knowledge sync)
6. Document everything thoroughly

SUCCESS CRITERION: All 109 loops verified working and tested, plus a working autonomy decision engine (AUTO / GATED / REFUSE permission gates) tested on 3+ pilot agents in shadow mode for 2 weeks with zero production incidents.

BUDGET: $400-450 USD fixed price, RELEASED ONLY IN MILESTONES as each is independently verified by us (not on your say-so):
- M1 (30%): Loop audit complete, 50% of repairs done
- M2 (30%): All loops repaired, test suite passes 109/109
- M3 (20%): Autonomy engine designed, 3-4 core loops implemented and tested
- M4 (20%): Full documentation, all verification complete

NO MONEY RELEASES UNTIL THE MILESTONE'S SUCCESS CRITERIA ARE INDEPENDENTLY VERIFIED.

START: ASAP, ideally 2026-08-30
COMMITMENT: Full-time, 2 weeks (~160 hours)

WE PROVIDE: Read access to codebase, live Linux test environment, all reference docs (loop specs, agent architecture, tooling).
YOU PROVIDE: DevOps/infrastructure automation expertise, strong bash scripting, systematic testing discipline, clear async written communication (no video calls required).

TO APPLY, INCLUDE:
1. Your DevOps/infrastructure automation background (5 sentences)
2. An example of a substantial bash automation project you've built (100+ lines, with error handling/logging)
3. Your understanding of what "loop repair + testing" means for this kind of system (2-3 sentences)
4. Confirmation you can commit full-time starting 2026-08-30 for 2 weeks

Full detailed technical specification (with exact deliverables, acceptance criteria, and constraints) will be shared with shortlisted candidates before award. Selected freelancer will get read access to the actual codebase and loop specs to review before formally starting.

Qualification requirements: 3+ years DevOps/infrastructure experience, strong bash (not just basic scripting), understanding of cron/systemd/scheduling, comfortable with async-only communication. Arabic language is a plus but not required.