Automated README summary
Automatically summarized from the README on 24 Oct 2026; not a run by jevbooks. Run instructions and author-reported outcomes still need verification.
What it is
Edward is an independent runtime supervisor for unattended coding agents (CI jobs, overnight runs, scheduled batches). It watches the event stream across turns, detects runaway patterns like loops, budget burn, and dangerous commands, and intervenes with PAUSE, CANCEL, or BLOCK, keeping a signed decision trail.
How it uses Jev
With EDWARD_SCORER_BACKEND=jev, all probes in a probe battery are batched into one calibrated Jev call over the cross-turn trajectory state. The verdict decides continue, pause, or escalate; low-confidence verdicts route to a human, while deterministic rules still handle dangerous-command blocking, budget caps, and the Ed25519-signed receipt chain.
Primitives:choice
Technique worth stealing
Batch all trajectory probes into one calibrated call, then route low-confidence verdicts to a human while deterministic rules keep hard safety guarantees.
Try it
pipx install edward-guard; edward demo --live-scorer --offline; set EDWARD_SCORER_BACKEND=jev and TYPESAFE_API_KEY, then edward wrap -- pi "task".
Model judgments from README excerpts
Project signals use repository metadata, an existing summary and the first 700 README characters, not the full project. “Not found” means not detected in that material; it does not establish absence. These are the original judgments with recorded models and dates, not source audits or run results.
Models & evaluation dates
- Inclusion gate
jev-1.14.0- Project signals
jev-1.14.0- Category
jev-1.14.0- Pattern tags
jev-1.14.0
Evidence
These signals use repository metadata, an existing summary and at most 700 cleaned README characters. ≥ 0.60 means reported in that excerpt; ≤ 0.40 means not found there. The middle band is inconclusive. None is a source-code audit or a test result. The runnable-tool question excludes research experiments; the recommendation question asks for a learning shortlist, not a quality verdict.
- Jev is core to the projectFound in material0.79
- Shows a System One patternFound in material0.80
- Describes handling uncertaintyFound in material0.90
- Reports measured numbersnot found in the excerpt0.05
- Is a runnable toolFound in material0.95
- Top-10 learning shortlist signalnot found in the excerpt0.26
- Is a replica of the modelnot found in the excerpt0.10
- Solves a common problem (0–2)score on a 0–2 scale1.83 / 2
- About JevFound in material0.90
no-judgment band (0.40–0.60)
no-judgment band (0.40–0.60)
The Jev model and date for each judgment stage are listed above. Card written by DeepSeek V4.1 Flash from the README on 24 Oct 2026.