[Senate] CI: Ambitious quest task generator — xhigh-effort LLM-driven strategic task creation open analysis:9 coding:7 instruction_following:9 reasoning:10 safety:8

Goal

Maintain a steady supply of ambitious, high-strategic-value tasks in the
SciDEX queue by running deep LLM analysis every 30 minutes. The work product
is new tasks (and priority adjustments on existing tasks) — never
boilerplate gap-counting. Output should advance core SciDEX capabilities
and accelerate cures and neuroscience, not just chip at row counts.

This spec deliberately replaces scripts/quest_engine.py and the prior quest-engine-ci.md template-driven approach, both of which produced
filler one-shots that mirror existing recurring CI work without adding
strategic leverage. The retired approach is documented for context only —
do not run it.

> ## Continuous-process anchor
>
> This spec instantiates theme S7 in docs/design/retired_scripts_patterns.md
> ("Quest task generation"). Every principle in the **"Design principles for
> continuous processes"** section is load-bearing. In particular:
>
> - LLMs for semantic judgment; rules for syntactic validation only
> - Gap-predicate driven, not calendar-driven
> - Idempotent + version-stamped + observable
> - No hardcoded entity lists, keyword lists, or canonical-name tables
> - Bounded batch (≤ 5 tasks per run unless deficit is larger)
> - Progressive improvement via outcome-feedback loop

The invariant this task enforces

At all times, SciDEX should have ≥ 5 open non-CI tasks at priority ≥ 90
that would, if completed, materially advance one of:

Core capability: scoring, debate, market mechanisms, KG quality,

tool reliability, governance, agent orchestration

Scientific output: hypothesis quality, evidence chains, falsifiable

predictions, experiment proposals, target validation for neurodegeneration

System value: SciDEX's role as a machine for prioritizing,

organizing, synthesizing, inventing, funding, and rewarding science

Gap predicate (run this first — if false, mostly-no-op):

SELECT COUNT(*) FROM tasks
WHERE project_id = (SELECT id FROM projects WHERE name='SciDEX')
  AND status IN ('open','available')
  AND priority >= 90
  AND task_type IN ('one_shot','iterative')
  AND title NOT LIKE '%CI:%'
  AND title NOT LIKE '%[Watchdog]%';

If COUNT() >= 5: skip generation. Run the priority audit* step only
(below). If COUNT() < 5: generate (5 - COUNT()) to min(10, deficit*2)
new tasks at priority ≥ 90.

What "ambitious" means here (the bar)

A task qualifies as ambitious if at least two of:

It targets a capability SciDEX doesn't have yet, or one whose current

implementation is a known-thin scaffold (not a row-count backfill).

It would, on completion, change the value of one of SciDEX's measurable

outputs (debate quality, market signal, hypothesis novelty, KG density,
reproducibility, time-to-first-experiment, agent throughput).

It frames a scientific question the system itself should answer

(cross-disease mechanistic synthesis, novel target proposal with
falsifiable prediction, paper-claim contradiction surface, etc.).

It builds a feedback loop / meta-mechanism (rubric versioning, tournament

driver, reward-eligibility check, calibration meta-job).

It improves cross-layer integration (Atlas ↔ Exchange, Agora ↔ Senate,

Forge ↔ Atlas) where today the layers are loosely coupled.

A task is filler (do not create) if it is any of:

• "Backfill column X for N rows" where a recurring CI task already covers

this (see §"Recurring tasks already covering common gaps" below).

• "Add references to N wiki pages" / "score N hypotheses" / "extract claims

from N papers" — these mirror existing every-6h drivers; the right
intervention is unsticking the driver, not creating a one-off chip.

• Anything whose acceptance is "process N rows" without naming a capability

improvement that wouldn't happen otherwise.

• Anything that's a wrapper around a Senate dedup/cleanup activity that

forces destructive action regardless of judgment.

What the agent does each cycle (xhigh effort)

You are running with --effort xhigh and should think hard before
committing. A single well-framed ambitious task is worth more than
five filler chips. Spend most of the cycle reading and synthesizing.

Phase 1 — State snapshot (read-only, ~5 min)

Gather and summarize:

Queue state via Orchestra MCP (list_tasks with project=SciDEX):

- Total open one-shot/iterative count
- Count at priority ≥ 90, ≥ 95, ≥ 99 (excluding CI: and Watchdog tasks)
- Top 25 highest-priority open tasks: id, title, priority, age, quest_id
- Top 10 oldest open tasks (for staleness review)
- Last 30 completed tasks (last 24h): id, title, layer, completion_summary

Recent run outcomes (last 24-48h):

- How many tasks completed vs abandoned vs still-running
- Which quests/layers shipped real PRs (check pr_links_json,
commit_links_json, merge_verified_at)
- Which agents/models were most productive (group assigned_worker)

Active quests (Orchestra quests table):

- For each active quest: name, layer, priority, current open-task count
- Read the quest's spec file under docs/planning/specs/quest_<layer>_spec.md
and any quest_<layer>_<topic>_spec.md companions
- Read the five mission quest specs (Q-DSC, Q-OPENQ, Q-LIVE, Q-PROP, Q-PERC)
when relevant to detect cross-quest leverage

Recurring CI health (critical — distinguishes filler from ambitious):

SELECT id, title, frequency, last_completed_at,
          (julianday('now') - julianday(last_completed_at)) AS days_stale
   FROM tasks
   WHERE project_id=(SELECT id FROM projects WHERE name='SciDEX')
     AND task_type='recurring' AND status='open'
     AND priority >= 90
   ORDER BY days_stale DESC NULLS FIRST LIMIT 30;

- If a recurring driver is stale > 24h, **a one-off chip at the same gap
is filler.** Either propose unsticking the driver as the new task, or
skip the gap entirely.

SciDEX world-model state via PostgreSQL (scidex.core.database).

Polymorphic via information_schema — never hardcode column lists.
Useful signals:
- Hypothesis: count by status, score-completeness distribution, novelty
histogram, evidence-link density
- Papers: ingestion rate, abstract/fulltext/figures coverage, claims-extracted
- Wiki: page count by content_md length percentile, refs_json density,
KG-linkage rate, mermaid-diagram coverage
- Markets: volume distribution, stale-resolution backlog, allocator activity
- Debates: sessions per hypothesis, scored-vs-unscored, recency
- KG: edges by type, low-confidence edge count, orphan node count
- Forge tools: tool_call success rate by tool, last_used_at staleness
- Senate: belief-snapshot recency, dividend distribution lag, contribution
credit pipeline depth, quality-gate failure rate

Site state (optional, when LLM judges relevant):

- curl -s http://127.0.0.1:8000/api/... for surface checks
- Read data/scidex-artifacts/ ToC (don't load everything)
- git log --since="24 hours ago" --format="%s%n%b" to see what landed

Phase 2 — Synthesis (xhigh effort think)

With state in hand, ask yourself the hard questions. Each answer is
candidate task material:

Strategic gaps (priority 95-99 candidates):

• What capability would change SciDEX's value proposition if shipped this

week? (Examples that would qualify: a working tournament driver that ranks
hypotheses by predictive accuracy; a reward-eligibility audit that closes
the contributor incentive loop; a cross-disease synthesis surface; a
literature-citation verifier replacing the hallucinated [PMID:NNNN]
markers; an agent-replay harness so failed tasks don't re-fail identically.)

• What scientific question is the system uniquely positioned to answer

this week? (e.g. "Generate 5 falsifiable mechanistic hypotheses connecting
microglial senescence to TDP-43 proteinopathy, with PubMed-cited evidence
for each, predictions, and proposed knockout experiments.")

• Which layers have the least cross-talk and what bridge would help most?

Capability gaps (priority 90-94 candidates):

• Which thin-scaffold capability is closest to ready-for-real-use? (e.g. an

experiment-proposal generator that has a working prompt but no UI surface.)

• Which feedback loop is open-loop today and would close with one task?

(e.g. debate quality → hypothesis prior; agent contribution → reward.)

Health gaps (priority 90-92, only if no recurring CI covers it):

• Which stuck recurring driver is the most leveraged to unstick? Frame as

"diagnose + fix + verify resumed throughput" — not "do the work the driver
was supposed to do."

Phase 3 — Bounded creation

For each new task:

Title: [Layer] <verb> <object> <specifics> — concrete, scannable.

Description: includes (a) why this matters (cite the strategic gap

from Phase 2), (b) what success looks like in measurable terms,
(c) what the agent should read first, (d) what not to do.

Priority: 90-99 by leverage. Only use ≥ 95 for genuine

capability-shift work; reserve 99 for "this changes the whole system."

Layer / quest_id: route via existing layer quests when fit; create

no new quests.

Spec_path: prefer existing quest layer specs (quest_<layer>_spec.md).

For genuinely novel ambitious work, write a new spec file under
docs/planning/specs/ describing the work in detail; commit it in the
same worktree before creating the task; reference its path in spec_path.

Task type: default iterative with max_iterations=15 for deep

work; use one_shot only for clearly-bounded single-PR work.

Tags: include quest-engine, layer name, and a topical tag.

Payload requirements: set realistic skill floors (reasoning,

analysis, coding, creativity, safety).

Cap creation at max(5, deficit + 2) per cycle. Never exceed 10.

Dedup: use the MCP create_task server-side check; additionally do a
fuzzy similarity ≥ 0.7 check against last 60 days of titles (open + done)
before submitting.

Phase 4 — Priority audit (every cycle)

Independent of creation, audit existing open task priorities:

• Overdue + low priority (e.g. 60d old at priority 50): demote to 30 or

close as stale (with rationale in completion_summary).

• Underprioritized strategic work: open task whose description references

a Phase 2 strategic theme but priority < 85 → bump to 85-90.

• Overprioritized filler: 95-priority task whose work is also covered

by an active every-6h recurring driver → demote to 70.

• Surface contention: if 3+ tasks are racing on the same files, lower

the priority of all but the most strategic.

Make ≤ 5 priority adjustments per cycle. Each adjustment: log the before/after
and the rationale in this spec's Work Log section.

Phase 5 — Cycle log

Append a Work Log entry with:

• UTC timestamp + worker id
• Phase 1 snapshot summary (≤ 10 lines)
• Phase 2 strategic gap synthesis (≤ 10 lines)
• Tasks created (id, title, priority, rationale 1-line each)
• Priority adjustments (id, before → after, rationale)
• Stuck recurring drivers flagged (id, days_stale, suggested intervention)
• Cycle duration

Recurring tasks already covering common gaps

Do not create one-shot duplicates of these (snapshot 2026-04-28; refresh
each cycle by querying tasks WHERE task_type='recurring' AND status='open'):

Gap	Recurring driver	Frequency
Hypothesis score updates from new debates	[Exchange] CI: Update hypothesis scores from new debate rounds	every-2h
Belief snapshots	[Economics] CI: Snapshot hypothesis prices for price history	every-2h
PubMed evidence backfill	[Exchange] CI: Backfill evidence_for/evidence_against with PubMed citations	every-6h
Discovery dividends	[Exchange] Discovery dividend backprop credit (driver #14)	every-6h
Squad dividend multiplier	[Exchange] Squad-member dividend multiplier on backprop (driver #23)	every-6h
Reward emission	[Exchange] Reward emission for contributions (driver #5)	every-6h
Token bounty issuance	[Exchange] Token bounty issuance for open work (driver #4)	every-6h
Funding allocator	[Exchange] Funding allocator activation (driver #10)	every-6h
Quadratic funding	[Exchange] Quadratic funding allocator (driver #15)	every-6h
Wiki citation enrichment	[Atlas] Wiki citation enrichment — add inline citations	every-6h
Wiki ↔ KG cross-linking	[Atlas] CI: Cross-link new wiki pages to KG entities	every-6h
Wiki mermaid regen	[Atlas] Wiki mermaid LLM regen	every-6h
Paper figures	[Atlas] Extract and reference figures from scientific papers	every-2h
Paper abstracts	[Forge] Reduce PubMed metadata backlog for papers missing abstracts	every-6h
Notebook coverage	[Artifacts] CI: Verify notebook coverage	every-12h
Stub notebook regen	[Artifacts] Audit all 67 stub notebooks	every-6h
Debate sessions	[Agora] CI: Trigger debates for analyses with 0 debate sessions	every-24h
Debate quality scoring	[Agora] CI: Run debate quality scoring on new/unscored sessions	every-6h
Counter-argument bounties	[Agora] Counter-argument bounty market (driver #7)	every-6h
Squad enrollment	[Agora] Squad open enrollment & recruitment (driver #21)	every-6h
Tool call failure triage	[Forge] CI: Test all scientific tools for availability	daily
World-model improvements	[Senate] World-model improvement detector (driver #13)	every-6h
Agent contribution credit	[Senate] Agent contribution credit pipeline (driver #11)	every-6h
Abandoned-run watchdog	[Senate] CI: Abandoned-run watchdog	every-1h
Strategic engine guardian	[Senate] Strategic engine guardian	every-15-min
Orchestra operator watchdog	[Senate] Orchestra operator watchdog and self-repair loop	every-2h

If a row-count gap matches one of these recurring drivers AND that driver
has run in the last 24h, do not generate a one-off duplicate. If the driver
is stale, prefer "unstick the driver" framing over "do the chipping."

Critical constraints

• Pri 99 every-30-min: this spec must run regardless of queue depth

(priority audit always happens; creation only when invariant violated).

• xhigh effort: agents picking this up should run with maximum

reasoning effort. The supervisor must route to a slot capable of
reasoning >= 10. (Note: per auth._ensure_claude_launch_flags, the
--effort xhigh flag is auto-repaired on acquire; non-interactive launch
is safe.)

• Provider any: claude or codex; both are acceptable. Codex is the

default for cycles requiring deep code reading.

• Idempotent: re-running this task immediately should be a no-op

(queue invariant satisfied → no creation; priority audit converges).

• No destructive action: this task only creates tasks and adjusts

priorities. It never deletes or merges artifacts. Never close another
task's lifecycle.

• Bounded blast radius: ≤ 10 task creations per cycle, ≤ 5 priority

adjustments per cycle. Anything larger needs operator gate.

• Worktree discipline: any new spec files must be written in a

worktree (per SciDEX guard-main-writes hook), committed, and PR'd.
Do not write to the main checkout.

• No new quests: route to existing quests; if no quest fits, route to

the most relevant layer quest (qe-<layer> family).

Acceptance criteria

Phase 1 before any creation Phase 5 every cycle's log, regardless of whether intervention happened

Dependencies

• /home/ubuntu/scidex/docs/planning/specs/ — quest specs to read
• /home/ubuntu/scidex/docs/design/retired_scripts_patterns.md — design principles
• /home/ubuntu/Orchestra/orchestra.db — task queue (read via MCP, write via MCP)
• scidex.core.database.get_db_readonly — SciDEX PostgreSQL access
• Orchestra MCP tools: list_tasks, create_task, update_task, list_quests

Dependents

• The agent fleet's productivity at any given moment is bounded by the

ambition of the queue this task generates.

• Strategic engine guardian (7a9c642b) reopens this if it gets blocked.
• Watchdog (433698d3 family) creates [Watchdog] Fix: tasks when this

abandons too often — those should be diagnosed, not silenced.

Work Log

2026-04-28 — Bootstrap (rewrite from quest-engine-ci.md)

Rewrote spec to replace the queue-depth-only template generator with an
xhigh-effort LLM-driven strategic generator. Prior approach (quest_engine.py
+ quest-engine-ci.md) produced filler one-shots that mirrored existing
every-6h recurring drivers without adding strategic leverage. New spec:

• Gap predicate is "≥5 non-CI priority-≥90 tasks", not "queue size < 50"
• Every cycle reads recurring CI health and refuses to chip at gaps that

active drivers already own

• Phases mandate state snapshot → strategic synthesis → bounded creation

→ priority audit → log

• "Ambitious" bar: capability work, scientific output, value-prop shift
• Hard prohibition on row-count backfill duplicates of existing drivers

Operators: see [Watchdog] Fix: tasks 433698d3, ff42baa0, 2fb403f8, 1fd42f17 for the abandonment history of the prior implementation.

Cycle 1 — 2026-04-28T09:11Z (first agent run on the new spec)

World model snapshot (Phase 1):

Metric	Value	Trend
Hypotheses	1,873	+1,117 in 7d
All in proposed lifecycle	1,873 / 1,873	CRITICAL
Debated hypotheses	1,866	—
Avg composite score	0.561	—
Top composite score	0.96	—
Debates (7d)	501	active
Knowledge gaps	3,545	3,153 open
Gaps resolved	12 / 3,545	0.34%
Wiki pages	17,662	—
KG edges	697,224	—
Papers	29,503	—
Analyses	470	—
Benchmarks	0	CRITICAL
ML models (artifacts)	9	low
Debate sessions	835 total	—

Queue health: 56 open non-CI tasks at priority ≥ 90 (gap predicate met),
97 recurring drivers active, only 3 one-shot tasks open (low for a
platform with this activity level).

Strategic gaps synthesized (Phase 2):

Hypothesis lifecycle frozen — every hypothesis stuck in proposed;

no promotion engine exists. World model can never improve.

Gap resolution rate 0.34% — 3,545 gaps, 12 resolved; gap tracker is

noise, not signal.

Zero benchmarks — Forge has no computational benchmarks; platform

lacks predictive-validity demonstration.

Debate transcripts unused for causal KG — 835 debates contain

mechanistic A→B→C reasoning never extracted as KG edges.

No cross-disease mechanism mining — debates siloed by disease;

no AD/PD/ALS/FTD analogy miner.

Top hypotheses lack computational validation — top 25 at

composite_score ≥ 0.88 debated but not computationally validated.

Ambitious tasks created (Phase 3):

ID	Title	Priority
7db6be96	[Agora] Hypothesis lifecycle promotion engine — debate-to-validated pipeline	94
c17abaf7	[Forge] Computational validation of top 25 hypotheses — enrichment + expression analyses	93
f4f7b129	[Atlas] Gap closure pipeline — match 500 open gaps to evidence and resolve	92
72f50712	[Agora] Debate transcript causal claim extractor — mine 835 debates for mechanistic KG edges	91
1186a9ab	[Forge] Seed neurodegeneration ML benchmark registry — 5 prediction tasks with answer keys	91
1b1ebf23	[Agora] Cross-disease mechanism analogy miner — transfer AD/PD/ALS/FTD insights	90

All six pass the "ambitious" bar (capability work, scientific-output,
or value-prop shift — not row-count backfill). None duplicate any of
the 26 recurring drivers in the inline overlap table.

Result observed at 10:16 UTC: Task 72f50712 (debate causal extractor)
already at 2,020 KG edges from 80/200 sessions — exceeded its 2,000-edge
target. The new generator is producing tasks that actually move the needle.

Cycle 2 — 2026-04-28T18:35Z (worker claude-auto:41)

Phase 1 — World model snapshot:

Metric	Value	Change from C1
Hypotheses	1,878	+5
Proposed	1,216 / 1,878	DOWN from 1,873 (promotion engine working)
Promoted	216	new status
Active	133	new status
Avg composite score	0.561	stable
Hypotheses score > 0.8	77	—
Knowledge gaps	3,545	same; 3,153 open; 12 resolved (0.34%)
Wiki pages	17,662	stable
KG edges (PostgreSQL)	2,316	up from ~800 (debate extractor impact visible)
Papers	29,503	stable
Analyses	471	+1
Debate sessions	841	+6
Artifacts	51,832	—

Gap predicate result: 0 non-CI one_shot/iterative tasks at priority ≥90 (VIOLATED; need ≥5).
Queue: 120 recurring drivers, 3 iterative tasks (all at priority 83–86), 0 one_shot.

Stale recurring drivers flagged (>24h since last_completed_at):

Driver	Priority	Days Stale	Note
Database integrity check	98	26d	CRITICAL systemic failure
Hypothesis score update	96	26d	Data quality frozen
Trigger debates	94	26d	Agora input frozen
Evidence backfill	94	26d	Evidence layer frozen
Debate quality scoring	93	26d	Quality signals frozen
King of the Hill tournament	97	22d	Core value prop broken
Evolve economics	97	24d	Markets frozen

Phase 2 — Strategic gaps synthesized:

Gap predicate violated: 0 eligible tasks (down from 56 in Cycle 1 — those were miscounted as recurring tasks). Immediate creation needed.

8+ recurring drivers all stale since ~2026-04-02: shared stale date = systemic root cause, not individual driver bugs. Fixing it at the source is the highest-leverage intervention.

King of the Hill tournament 22d stale: Platform's core hypothesis-ranking mechanism is non-functional. Prediction markets are frozen.

KG still sparse: 2,316 edges despite ~760 unprocessed debate sessions that yield ~19 edges each. Task 72f50712 proved the extraction works (1,498 edges from 80 sessions on 2026-04-16); completing it = ~7x KG density.

Gap resolution rate 0.34% unchanged: No resolution engine exists. Building one converts 3,153 open gaps from noise to progress signal.

Cross-disease synthesis absent: 841 debates span AD/PD/ALS/FTD but no synthesis surface extracts shared pathways — unique scientific value only SciDEX can produce at scale.

Top hypotheses lack structured output: 77 hypotheses at composite_score > 0.8 exist as DB rows; no research brief format exists for researchers to act on.

Phase 3 — Tasks created:

ID	Title	Priority	Rationale
565ac01b	[Senate] Diagnose and fix 26d+ stale recurring CI drivers	96	8+ drivers shared stale date = systemic root cause; fix the engine not the symptoms
a3018033	[Arenas] Fix King of the Hill tournament driver	96	Core hypothesis-ranking mechanism broken 22d; prediction markets non-functional
31eeae8d	[Agora] Gap resolution engine — auto-close gaps	95	3,153 open gaps, 12 resolved; build the resolution pipeline, not one-off chips
8ebfce57	[Atlas] Complete KG extraction from ~760 debate sessions	94	2,316 edges; ~7x density possible; 72f50712 proved approach works
ffd81f3a	[Agora] Cross-disease mechanism synthesizer	93	841 siloed debates; shared AD/PD/ALS/FTD mechanisms = unique scientific value
33dca458	[Agora] Research briefs for top 25 hypotheses	92	77 high-score hypotheses as bare DB rows; structured reports = platform's core deliverable

All 6 pass the ambitious bar (≥2 of: capability, measurable output change, scientific question, feedback loop, cross-layer integration). None duplicate active recurring CI work.

Phase 4 — Priority adjustments:

Task	Before	After	Rationale
66c83cdc [Forge] Benchmark answer-key migration to dataset registry	89	93	Strategic infra: benchmarks still zero, this is the pipeline to fix it
55e3ea08 [Atlas] Reduce wiki-to-KG linking backlog	82	50	Covered by recurring [Atlas] CI: Cross-link new wiki pages to KG entities; demoted until driver is unstuck
2f7e1600 [Agora] Add counter-evidence reviews to 10 hypotheses	86	60	Mirrors stale [Exchange] CI: Backfill evidence_for/evidence_against driver; new stale-driver task (565ac01b) will unstick the driver

Cycle outcome: Gap predicate was 0 → will be ≥6 after these tasks are claimed. Six spec files committed to worktree at SHA 90904c4f4. Stale driver diagnosis (565ac01b) is the highest-leverage task: fixing it would unblock 8+ drivers and restore data quality across all layers.

Cycle 3 — 2026-04-28T22:25Z (worker claude-auto, task 80ffb77b)

Phase 1 — World model snapshot:

Metric	Value	Change from C2
Hypotheses	1,967	+89
Proposed	1,277 / 1,967	stable
Promoted	221	+5 from C2
Active	156	+23
Validated	0	CRITICAL — unchanged
Avg composite score	0.564	+0.003
Hypotheses score ≥ 0.8	88	+11
Knowledge gaps	3,545	stable
Open gaps	2,635	down from 3,153 (gap resolution engine working)
Resolved gaps	529	up from 12 — massive improvement
KG edges	2,316	same total; 518 added last 24h (extraction ongoing)
Debate sessions	849	+8 from C2
Papers	29,560	+57
Wiki pages	17,667	+5
Experiments	724	new (2bad68f6 task shipped)
Hypothesis predictions	3,729	new (bulk predictions exist)
Benchmarks	6	new (seeded by 1186a9ab)
Benchmark submissions	1	only 1 submission for 6 benchmarks
Prediction markets	84	stable
Cross-disease analogies	18	new (ffd81f3a shipped)

Gap predicate result: 3 (< 5, gap predicate VIOLATED; deficit = 2, creating 3 tasks)
Queue: 120 recurring drivers, 3 iterative tasks (d3d8cace @95, 8ebfce57 @94, 03094ddf @92).

Stale recurring drivers note: All 120 recurring tasks show last_completed_at = NULL
(999d stale), which is a known tracking artifact — the field isn't being populated, not a
real staleness signal. No action taken on this since a one-shot duplicate would be filler.

Recent merged work (last 48h, notable):

• 8180d43b2 — KG edge extraction (8ebfce57) merged; still iterating
• b3668c46b — KG causal extraction (d3d8cace) merged; still iterating
• a840eae9e — Cross-disease mechanism synthesizer (ffd81f3a) — 18 analogies
• 0bb597081, bf472125e — Research briefs (33dca458) — merged
• 8ea8cc96b — Experiment proposal generator (2bad68f6) — 724 experiments created
• df1ad15de — Gap resolution engine (31eeae8d) — 529 gaps resolved
• 473783fee — Gap closure pipeline (f4f7b129) — batch gap closure

Phase 2 — Strategic gaps synthesized:

0 validated hypotheses (CRITICAL): 221 promoted, 88 composite_score≥0.8, 3,729

predictions, 724 experiments — yet no hypothesis is "validated". The lifecycle gate is absent.
This is the most important scientific output gap on the platform.

6 benchmarks with 1 submission: Seeding task (1186a9ab) built the registry but not

the evaluator. Hypotheses have no benchmark scores — they can't be compared quantitatively.

Forge↔Exchange bridge missing: 724 experiments and 84 prediction markets are disconnected.

No pathway from experiment→market exists. Researchers can't trade on experimental predictions.

KG edge growth active but slow: 518 edges added in 24h but total still 2,316 with 849

debates to mine. Tasks d3d8cace and 8ebfce57 are still running iteratively; no new task needed.

All recurring drivers show NULL last_completed_at: Tracking bug, not real staleness.

Not actionable via new one-shots.

Phase 3 — Tasks created:

ID	Title	Priority	Rationale
(pending)	[Agora] Hypothesis validation gate	95	0 validated out of 88 eligible; lifecycle gate is absent; closes the core science loop
(pending)	[Forge] Benchmark evaluation harness	92	6 benchmarks, 1 submission; evaluation harness missing; turns debates into testable predictions
(pending)	[Exchange] Experiment-to-prediction-market bridge	91	724 experiments + 84 markets disconnected; Forge↔Exchange cross-layer bridge

All 3 pass the ambitious bar (≥2 of: capability, measurable output change, scientific question,
feedback loop, cross-layer integration). None duplicate active recurring CI work.

Phase 4 — Priority adjustments:

No priority adjustments this cycle. The 3 open non-recurring tasks (d3d8cace @95, 8ebfce57 @94,
03094ddf @92) are appropriately prioritized — all are active iterative tasks producing real work.
Recurring task staleness values (all 999d) are a tracking artifact, not real stale dates.

Cycle outcome: Gap predicate 3 → will be ≥6 after these 3 tasks are claimed. Three new
spec files committed. Top priority: validation gate (0 validated hypotheses is the starkest gap
visible in the world model).

Cycle 4 — 2026-04-29T03:05Z (worker claude-auto, task 80ffb77b)

Phase 1 — World model snapshot:

Metric	Value	Change from C3
Hypotheses	1,974	+7
Proposed	1,284	+7
Promoted	221	stable
Active	156	stable
Validated status	does not exist in schema	CRITICAL
High-score (≥0.8)	88	stable
Avg composite score	0.564	stable
KG gaps	3,545	stable
Open gaps	2,635	stable
Resolved gaps	529	stable
Papers	29,566	+6
Wiki pages	17,682	+15
KG edges (kg_edges)	2,366	+50
Causal edges (causal_edges)	19,753	HUGE — wiki-extracted
Debate sessions	850	+1
Experiments	724	stable
Benchmarks	6	stable
Benchmark submissions	227	was 1 in C3 — harness working!
Predictions pending	3,719	+12
Predictions confirmed	9	NEW
Predictions falsified	5	NEW
Prediction markets	139	+120 in 24h (experiment bridge active)
Resolved markets	23	+23 in 24h (resolution engine active)
Convergence reports	61	NEW — previously unseen
Cross-disease analogies	18	stable

Gap predicate result: 1 (< 5, VIOLATED; only 8ebfce57 qualifies; deficit = 4)
Queue: 118 recurring, 1 iterative (8ebfce57 @94), 0 one-shot non-CI.

Stale recurring drivers: All 118 recurring tasks show last_completed_at = NULL
(999d stale) — confirmed tracking artifact, not real staleness. No action.

Notable recent merges:

• 2a40d719a — Atlas mechanism consensus map (1,974 hypotheses classified by pathway)
• 136cc0f22 — Forge benchmark evaluation harness (shipped 226 new submissions)
• 56726030f — Exchange experiment-to-prediction-market bridge (120 markets in 24h)
• ca6dfe65a — Hypothesis validation gate (merged but 'validated' status doesn't exist in schema)
• 535106909 — Prediction market resolution engine (23 resolved in 24h)
• edb667451 — Hypothesis prediction contradiction detector

Phase 2 — Strategic gaps synthesized:

'validated' status missing from schema (CRITICAL): validation gate (ca573a56) shipped

but hypotheses.status has no 'validated' option — status enum: proposed/promoted/active/
debated/archived/open/superseded. 88 hypotheses at composite_score ≥ 0.8 have no
scientific endpoint. Platform cannot deliver validated science.

19K causal edges siloed: causal_edges table (19,753, wiki-extracted, free-text entities,

mechanism_description, evidence_pmids) is disconnected from kg_edges (2,366, typed entity refs).
10x KG density increase possible via entity resolution bridge.

61 convergence reports unsurfaced: convergence_reports table has 61 rows but they're not

quality-scored, not UI-surfaced, not linked to hypotheses — potentially the platform's first
scientific synthesis output, invisible to users.

3,741 predictions 99.6% unevaluated: infrastructure exists (9 confirmed, 5 falsified found

manually), but systematic evaluation pipeline is absent. Evaluating predictions demonstrates
platform's predictive validity.

120 prediction markets in 24h without quality gate: experiment bridge is working but

creating noise. Markets need resolution criteria, non-50/50 priors, and liquidity > 0.

Phase 3 — Tasks created:

ID	Title	Priority	Rationale
0b9657cd	[Agora] Add 'validated' hypothesis lifecycle status	96	'validated' missing from schema; 88 eligible; closes scientific closure loop
12c461ae	[Atlas] Causal KG entity resolution	95	19K free-text edges → KG bridge; 10x density; Atlas↔Agora integration
712ca5de	[Atlas] Surface and score 61 convergence reports	93	First scientific synthesis output; 61 reports exist but invisible
2c4b95b0	[Agora] Falsifiable prediction evaluation pipeline	92	3,741 pending predictions; 0.37% evaluated; closes prediction feedback loop
41ee05e3	[Exchange] Prediction market quality gate	91	120 markets in 24h; quality gate missing; protects market signal integrity

All 5 pass the ambitious bar (≥2 of: capability, measurable output, scientific question,
feedback loop, cross-layer integration). None duplicate any of the 26 recurring drivers.

Phase 4 — Priority adjustments:

No adjustments this cycle. The 1 open non-recurring task (8ebfce57 [Atlas] KG edge
extraction @94) is appropriately prioritized and actively iterating. Recurring driver
staleness values are tracking artifacts. New tasks span P91-P96 with appropriate gradation.

Cycle outcome: Gap predicate was 1 → will be 6 (5 new + 8ebfce57) after tasks are claimed.
5 spec files committed (SHA 28ad1fe9d). Top priority: 0b9657cd — 'validated' status missing
from schema is the most critical scientific lifecycle gap; without it, the platform cannot
deliver its core promise (validated science). Secondary priority: 12c461ae causal KG bridge
(10x KG density increase with existing data).

Cycle 5 — 2026-04-29T03:55Z (worker claude-auto, task 80ffb77b)

Phase 1 — World model snapshot:

Metric	Value	Change from C4
Hypotheses	1,974	stable
Proposed	1,284	stable
Promoted	221	stable
Active	156	stable
Validated	0	CRITICAL — unchanged (validation task in flight)
High-score (≥0.8)	88	stable
Avg composite score	0.564	stable
Knowledge gaps	3,545	stable
Open gaps	2,635	stable
Resolved gaps	529	stable
Wiki pages	17,682	stable
KG edges	2,366	stable
Causal edges	19,753	stable
Papers	29,566	stable
Debate sessions	864	+14
Experiments	724	stable
Benchmarks	6	stable
Benchmark submissions	227	stable
Convergence reports	61	stable
Prediction markets	139 (80 active, 23 resolved, 36 cancelled)	stable
Tournaments	112 (95 complete, 16 open, 1 in_progress)	stable
Tournament matches	2,207	stable
Targets	185	stable
Target dossier rows	0	CRITICAL — new gap identified
Wiki quality scores	5,563	—
World model improvements	1,309	—

Gap predicate result: 1 (only 8ebfce57 open; Cycle 4 tasks 0b9657cd, 12c461ae, 712ca5de,
2c4b95b0 all moved to running status within 30min of creation — gap predicate counts open/available only, so running tasks don't qualify). Deficit = 4.

Running non-CI tasks at p≥90 (not counted by predicate, all in flight): 0b9657cd p=96 (validated lifecycle), 12c461ae p=95 (causal KG), d924270b p=94
(hypothesis advancement optimizer — new, created outside this generator), 712ca5de p=93
(convergence reports), 03094ddf p=92 (falsifiable predictions), 2c4b95b0 p=92
(prediction evaluation), 17f3b8e0 p=91 (mechanistic claim verifier).

Stale recurring drivers: All recurring tasks show last_completed_at = NULL (999d stale)
— confirmed tracking artifact from prior cycles, not real staleness. No action.

Phase 2 — Strategic gaps synthesized:

185 targets, 0 dossier rows (CRITICAL): The target_dossier table has zero rows

despite 185 neurodegeneration drug targets existing. No running task addresses this.
Targets are the translational endpoint — the entities drug discoverers act on.

Debate sessions as unused hypothesis source: 864 debate sessions contain structured

multi-agent reasoning (claims with agent confidence, skeptic concessions, consensus
points) that has never been mined for NEW hypothesis candidates. This is richer than
single-LLM generation — cross-agent consensus is a quality signal.

No hypothesis-target linkage: 1,974 hypotheses and 185 targets exist in parallel

with no systematic relevance scoring between them. unit_hypothesis_links exists but
population is unknown. Linking hypotheses to targets creates the drug discovery ranking.

Agent contribution loop unverified: agent_contributions, agent_reputation,

token_reward_events tables exist with recurring drivers, but the full pipeline from
task completion → reputation update → token emission has never been audited end-to-end.
A broken link means the incentive system is non-functional.

9 ML models vs 6 benchmarks, no leaderboard: The Forge↔Atlas measurement loop

(artifacts evaluated on benchmarks) has no summary leaderboard despite 227 submissions.

Phase 3 — Tasks created:

ID	Title	Priority	Rationale
fc9309fd	[Atlas] Neurodegeneration target dossier pipeline	96	185 targets, 0 dossiers; direct translational output; closes Atlas-to-drug-discovery loop
4ba968e0	[Agora] Debate-to-hypothesis synthesis engine	94	864 debate sessions unused as hypothesis source; multi-agent consensus = quality signal
14954567	[Atlas] Hypothesis-target relevance scoring	93	No hypothesis-target linkage; creates drug discovery priority ranking
3ff23dae	[Senate] Agent contribution feedback loop audit	91	Full pipeline from task→reputation→token never verified end-to-end
fb4e9333	[Forge] ML model vs benchmark head-to-head evaluation	90	9 models + 6 benchmarks + 227 submissions but no leaderboard

All 5 pass the ambitious bar (≥2 of: capability, measurable output change, scientific
question, feedback loop, cross-layer integration). None duplicate any of the 26 recurring
drivers or the 7 tasks currently running.

Phase 4 — Priority adjustments:

No adjustments this cycle. The 7 running non-CI tasks (p=91–96) are appropriately
prioritized for their strategic value. Recurring task staleness values are tracking
artifacts. No priority misalignment detected.

Cycle outcome: Gap predicate was 1 → will be ≥6 after new tasks are claimed.
5 spec files committed SHA 5b69529dd. Top priority: fc9309fd (target dossier pipeline
— 185 targets with 0 dossiers is the starkest new gap). Secondary: 4ba968e0
(debate-to-hypothesis synthesis — unused high-quality source for novel hypotheses).