Asked to stop the research loop from re-testing feature ideas it had already tested. Came back with an automatic have-we-seen-this-before filter plus a registry that cannot be silently corrupted, promoted without evidence, or edited behind your back.
| Quantity | Value | Meaning |
|---|---|---|
| Iterations closed | 11 of 11 | every manifest row marked done |
| Catalog covered | 44 rows | provenance hash and chained history backfilled onto all 44 |
| Novelty-engine self-tests | 8 of 8 on the live catalog | exact, case/space, alias, novel, abbreviation, sibling-distinctness |
| Detection funnel | 3 tiers; lexical bands >= 95 / 85-95 / < 85 | duplicate / human-review gray zone / new |
| Robustness dimensions implemented | 6 of 6 (D1-D6) plus structural validation and Tier-3 | full blueprint |
| Ecosystems surveyed | 8 | source material for the blueprint |
| Firing | What it did |
|---|---|
| 02 | Surveyed 8 mature ecosystems and turned them into one prioritized hardening plan |
| 03 | Operator caught a conflation of novelty and orthogonality; every artifact corrected |
| 04 | Three-tier duplicate-detection engine shipped, stdlib only, checked against the real catalog |
| 06 | Illegal stage jumps refused; every row stamped with a content-identity hash |
| 09 | A candidate can no longer be promoted on a stale or unprovenanced verdict |
| 11 | Abbreviation synonyms now flagged for review instead of slipping through as new |
Iteration 11's own recommendation: review and merge PR #535 and stop the loop. Remaining work is polish only — extend the abbreviation map as new families appear; an embedding-based semantic tier stays optional behind the same gate. The 44 existing ids were deliberately not re-minted because content identity already lives in spec_hash.
'Semantic Tier-3 (abbreviation expansion) — the last gap closed, blueprint complete.' All 11 manifest rows read DONE; the iter-11 page recommends closing the campaign.
Lifecycle status is process state, not a judgement of the findings. Results are stated as numbers with their uncertainty. Generated between CLOSING-RECONCILIATION markers — regenerate rather than edit.
opendeviationbar-py · three-axis probes · loop campaign
A self-paced loop that makes the candidate registry — the sign-in sheet that tracks every feature idea stage-by-stage — more robust, and turns the slow manual "is this already covered?" check into a fast, automatic, evidence-backed pre-filter. Each iteration below explains, in plain language, what has been grounded, what is next, and how it moves us toward finding new orthogonal features.
| # | iteration | status | headline |
|---|---|---|---|
| 01 | Grounding the registry + standing up the campaign | DONE | Mapped the whole candidate-tracking system end-to-end and stood up this loop's dashboard so every step is visible. |
| 02 | SOTA research synthesized into a robustness blueprint | DONE | Surveyed how 8 mature ecosystems keep a registry trustworthy, then combined them into one concrete, prioritized hardening plan for our ledger — now written into the shared-data folder. |
| 03 | Operator steer — Novelty ≠ Orthogonality (terminology locked) | DONE | The operator caught a conflation in iteration 02: novelty and orthogonality are two DISTINCT gates. Locked the definition and corrected every artifact that blurred them. |
| 04 | P0 shipped — the novelty gate is now runnable, tested code | DONE | Built the three-tier novelty dedup engine (the fast catalogue-uniqueness pre-filter) as a stdlib-only module and verified it against the real 44-candidate catalog — 8/8 checks pass. |
| 05 | Novelty gate wired into the ledger + fleet-safe writes (D2) | DONE | Adding a candidate now AUTO-runs the novelty check and records the result, and the ledger can no longer be corrupted or lose appends when multiple sessions write at once — verified end-to-end without touching the real ledger. |
| 06 | Lifecycle moves now policed (D3) + provenance backfilled onto all 44 rows | DONE | Stage changes are now checked against a legal transition table (illegal jumps are refused), and every existing candidate was stamped with a content-identity hash so the dedup system covers the whole catalog. |
| 07 | Promotion machine-enforced + the registry is now /registry:* commands | DONE | A candidate can no longer be promoted without evidence, and every registry capability is now a discoverable slash command with a one-glance health check — making the PR capability-complete per the Eon standard. |
| 08 | Tamper-evident, hash-chained history (P2/D6) | DONE | Every change to a candidate is now a hash-chained event — editing any past entry out-of-band breaks the chain and the registry detects it. Backfilled onto all 44 rows and verified. |
| 09 | Verdicts pinned to a data/code snapshot — evidence before promotion (P2/D5) | DONE | A candidate can no longer be promoted on a stale or unprovenanced verdict: each axis verdict is now bound to the exact data + code that produced it, and promotion requires that pin to be complete. |
| 10 | Collision-proof IDs + typo-proof validation + a README — blueprint substantially complete | DONE | New candidates now get a collision-proof id automatically, stray/typo fields are rejected, and a README ties the whole subsystem together. All six robustness dimensions are implemented. |
| 11 | Semantic Tier-3 (abbreviation expansion) — the last gap closed, blueprint complete | DONE | The novelty gate now also catches abbreviation synonyms (e.g. A-MF-DFA ≡ Asymmetric MF-DFA) — dependency-free, conservative, flag-for-review. Every blueprint dimension is now implemented. |