nsaspy/prolog-rlm

Fork 0

[ARADR][MACHINE-SPIRIT] 8 topics × 8 alternating research/design passes #387

New issue

Open

opened 2026-09-08 01:29:17 +00:00 by nsaspy · 1 comment

nsaspy commented

2026-09-08 01:29:17 +00:00

Owner

Parent/source research: #397 #398 #399 #400 #401 #402 #403 #404 #405
Canonical handoffs: #376-#387 #388 #392-#396 and downstream symbolic-memory Machine Spirit issues.

Mission

Re-realize the completed Cycle-1 Machine Spirit research as an ARADR adaptive research/design program with 8 passes per topic, strict phase separation, and dependency-aware auto mode.

This issue does not authorize Machine Spirit implementation. It normalizes/deepens research and design and produces implementation handoffs only after the evidence is complete.

Hard mode invariant

Each loop has exactly one mode:

RESEARCH or DESIGN

Never both.

If a DESIGN loop discovers missing evidence, record a research-gap node and defer it to the next RESEARCH loop. A RESEARCH loop may record design implications but may not select/approve canonical architecture.

Eight passes per topic

For each topic:

R1 foundational research
D1 foundational design
R2 adversarial/deep research
D2 hardened design
R3 composition/scale/interoperability research
D3 integrated design
R4 executable-reality/conformance research
D4 final v1 design + implementation handoff

Baseline total: 8 topics × 8 passes = 64 major passes (32 research, 32 design).

Each pass should stack the largest coherent slice possible; do not fragment into trivial micro-passes.

Topics / canonical seeds

MS1 Knowledge Representation — #398 #392
MS2 Semantic Compilation — #399 #393 #395 #396
MS3 Epistemics / Truth Maintenance — #400 #394
MS4 Procedural Intelligence / Expert Systems — #401 #376-#387
MS5 Durable World Model — #402 + symbolic-memory
MS6 Retrieval / Symbolic Attention — #403 #381 #396 #219
MS7 Bounded Self-Improvement — #404
MS8 Machine Spirit Synthesis / Proving Ground — #405 #397

ARADR expresses these as a dependency DAG; dependency excursions are allowed, but mode mixing is not.

Evidence/status boundary

Preserve the proving-ground law:

issue says implemented
!= code exists
!= test exists
!= test executed
!= deterministic evidence
!= Machine Spirit acceptance

Current Cycle-1 verdict remains:

research/design: complete
Machine Spirit executable acceptance: NOT MET

Research-completion gate

Do not declare v1 research complete just because all 64 passes ran. Every v1-required unresolved item must end as one of:

answered
operator decision required
implementation experiment required
benchmark required
explicitly deferred beyond v1

There must be no unresolved research required node necessary for v1.

Zara transition after research completion

Only after the Machine Spirit research-completion gate passes:

pause all Zara workers except one;
preserve paused workers' state/worktrees/findings/blockers for resume;
retain exactly one Zara implementation worker covering:
- Zara core/desktop/runtime
- Zara Android
- Zara plugins/plugin integration
- cross-surface Zara implementation
consume the existing dependency-ordered Zara backlog; do not create a new feature program.

Record the transition with a Machine Spirit research-completion receipt and the paused/retained worker state.

Current execution

MS1-R1 is now active. It is research-only and will persist findings/gaps without making design decisions.

Parent/source research: #397 #398 #399 #400 #401 #402 #403 #404 #405 Canonical handoffs: #376-#387 #388 #392-#396 and downstream symbolic-memory Machine Spirit issues. ## Mission Re-realize the completed Cycle-1 Machine Spirit research as an **ARADR adaptive research/design program** with **8 passes per topic**, strict phase separation, and dependency-aware auto mode. This issue does **not** authorize Machine Spirit implementation. It normalizes/deepens research and design and produces implementation handoffs only after the evidence is complete. ## Hard mode invariant Each loop has exactly one mode: ```text RESEARCH or DESIGN ``` Never both. If a DESIGN loop discovers missing evidence, record a research-gap node and defer it to the next RESEARCH loop. A RESEARCH loop may record design implications but may not select/approve canonical architecture. ## Eight passes per topic For each topic: ```text R1 foundational research D1 foundational design R2 adversarial/deep research D2 hardened design R3 composition/scale/interoperability research D3 integrated design R4 executable-reality/conformance research D4 final v1 design + implementation handoff ``` Baseline total: **8 topics × 8 passes = 64 major passes** (32 research, 32 design). Each pass should stack the largest coherent slice possible; do not fragment into trivial micro-passes. ## Topics / canonical seeds 1. **MS1 Knowledge Representation** — #398 #392 2. **MS2 Semantic Compilation** — #399 #393 #395 #396 3. **MS3 Epistemics / Truth Maintenance** — #400 #394 4. **MS4 Procedural Intelligence / Expert Systems** — #401 #376-#387 5. **MS5 Durable World Model** — #402 + symbolic-memory 6. **MS6 Retrieval / Symbolic Attention** — #403 #381 #396 #219 7. **MS7 Bounded Self-Improvement** — #404 8. **MS8 Machine Spirit Synthesis / Proving Ground** — #405 #397 ARADR expresses these as a dependency DAG; dependency excursions are allowed, but mode mixing is not. ## Evidence/status boundary Preserve the proving-ground law: ```text issue says implemented != code exists != test exists != test executed != deterministic evidence != Machine Spirit acceptance ``` Current Cycle-1 verdict remains: ```text research/design: complete Machine Spirit executable acceptance: NOT MET ``` ## Research-completion gate Do not declare v1 research complete just because all 64 passes ran. Every v1-required unresolved item must end as one of: ```text answered operator decision required implementation experiment required benchmark required explicitly deferred beyond v1 ``` There must be no unresolved `research required` node necessary for v1. ## Zara transition after research completion Only after the Machine Spirit **research-completion gate** passes: - pause all Zara workers except one; - preserve paused workers' state/worktrees/findings/blockers for resume; - retain exactly one Zara implementation worker covering: - Zara core/desktop/runtime - Zara Android - Zara plugins/plugin integration - cross-surface Zara implementation - consume the existing dependency-ordered Zara backlog; do not create a new feature program. Record the transition with a Machine Spirit research-completion receipt and the paused/retained worker state. ## Current execution **MS1-R1 is now active.** It is research-only and will persist findings/gaps without making design decisions.

nsaspy referenced this issue

2026-09-08 01:29:19 +00:00

[MACHINE-SPIRIT 4/8] Procedural intelligence and expert systems: production rules, blackboards, HTN planning, CBR, diagnosis and repair #397

nsaspy referenced this issue

2026-09-08 01:29:20 +00:00

[EPIC][cyber-harness] Generic Prolog-RLM cyber reasoning harness for Hackmode/Hackpert #409

nsaspy referenced this issue

2026-09-08 01:46:19 +00:00

[EPIC][P0] First-class Prolog expert system: registry, routing, specialists, and closed-loop Spec→Plan→Expert→Verify→Repair #422

nsaspy referenced this issue

2026-09-08 01:46:54 +00:00

[EPIC][experts] Closed-loop integration: Requirement→Knowledge→Plan→Experts→VERIFY→Repair→Review→completion #411

nsaspy referenced this issue

2026-09-08 01:46:54 +00:00

[EPIC][experts] Closed-loop integration: Requirement→Knowledge→Plan→Experts→VERIFY→Repair→Review→completion #411

nsaspy referenced this issue

2026-09-08 01:47:23 +00:00

[semantic-eval] Cross-domain semantic conformance corpus and meaning-preservation / hallucination-loss metrics #403

nsaspy referenced this issue

2026-09-08 01:47:46 +00:00

[MACHINE-SPIRIT 4/8] Procedural intelligence and expert systems: production rules, blackboards, HTN planning, CBR, diagnosis and repair #397

nsaspy referenced this issue

2026-09-08 01:47:47 +00:00

[MACHINE-SPIRIT 4/8] Procedural intelligence and expert systems: production rules, blackboards, HTN planning, CBR, diagnosis and repair #397

nsaspy referenced this issue

2026-09-08 01:47:59 +00:00

[MACHINE-SPIRIT 8/8] Adversarial synthesis: Machine Spirit v1 architecture, conformance suite, zero-LLM demonstrations, and implementation graph #393

nsaspy commented

2026-09-08 01:48:17 +00:00

Author

Owner

ARADR checkpoint — MS1-R1 COMPLETE → MS1-D1 ACTIVE

Completed research-only child: #412.

MS1-R1 RESEARCH: COMPLETE
canonical design changed: no
implementation changed: no
research gaps: MS1-G01..MS1-G07
current canonical semantic_compile/sem_context implementation: not present on main

The next semantic loop is now MS1-D1 DESIGN ONLY. It may consume #412 plus the prior #398/#392 corpus, but it may not perform new external research. Any missing evidence discovered during D1 will be recorded for MS1-R2 instead of being researched inside the design pass.

## ARADR checkpoint — MS1-R1 COMPLETE → MS1-D1 ACTIVE Completed research-only child: #412. ```text MS1-R1 RESEARCH: COMPLETE canonical design changed: no implementation changed: no research gaps: MS1-G01..MS1-G07 current canonical semantic_compile/sem_context implementation: not present on main ``` The next semantic loop is now **MS1-D1 DESIGN ONLY**. It may consume #412 plus the prior #398/#392 corpus, but it may not perform new external research. Any missing evidence discovered during D1 will be recorded for MS1-R2 instead of being researched inside the design pass.