GLP-1 theory and grant construction

Experimental contract

What this family tests

registered pilots and proposal-realization ladder

The current catalog registers this at family and run level rather than assigning separate study titles.

Book-to-code crosswalk

The highlighted Experiment cards covered here

1. Experiment: GLP1–PILOT–0: retrospective withdrawal-context transport.17.10.1 Treatment status as a missing context (GLP1–PILOT–0) · reported result or audit · direct family
2. Experiment: GLP1–PILOT–1: hidden-context recovery.17.10.2 Discovering the missing context (GLP1–PILOT–1) · reported result or audit · direct family
3. Experiment: GLP1–GRANT capstone: theory extension to research program.17.11 A grant proposal as a capstone artifact · reported result or audit · proposed capstone to reference family
4. Experiment: GLP1–GRANT–0: controlled program compilation.17.11.1 Structural compilation without semantic admission · reported result or audit · direct family
5. Experiment: GLP1–GRANT–1: semantic admission firewall.17.11.2 Semantic admission as a type system · reported negative or rejected · direct family
6. Experiment: GLP1–GRANT–2: typed IR and narrative round trip.17.11.3 Compiling prose from a typed intermediate program · reported negative or rejected · direct family
7. Experiment: GLP1–GRANT–3: cross-field scientific invariants.17.11.4 Scientific invariants across a research program · reported result or audit · direct family
8. Experiment: GLP1–GRANT–4: auditable proposal realization.17.11.5 Auditable rhetorical realization · reported negative or rejected · direct family
9. Experiment: GLP1–GRANT–4.1: sanitized auditable realization.17.11.6 Sanitized realization and surface separation · reported negative or rejected · direct family
10. Experiment: GLP1–GRANT–4.2: structure-first realization.17.11.7 Structure-first rhetorical realization · reported negative or rejected · direct family
11. Experiment: GLP1–GRANT–4.3: bounded realization.17.11.8 Local metric and semantic bounds · reported result or audit · direct family
12. Experiment: GLP1–GRANT–4.4: controlled paraphrase.17.11.9 Controlled paraphrase and evaluator failure · reported result or audit · direct family
13. Experiment: AGENTIC-CORPUS–0.17.13 Toward an integrated open-domain benchmark · proposed · proposed benchmark to reference family

Implementation map

From entry point to source surface

Start here

These concrete files are selected from the complete registered source surface to orient the reader.

• runner · staged-not-public — synthetic-creativity-archive:2026-08-20_glp1_grant0_program_compilation_registered/run.pyLocate this file in the curated companion packet; no public download is currently offered.
• runner · staged-not-public — synthetic-creativity-archive:2026-08-20_glp1_grant1_semantic_firewall_registered/run.pyLocate this file in the curated companion packet; no public download is currently offered.
• runner · staged-not-public — synthetic-creativity-archive:2026-08-20_glp1_grant2_typed_ir_roundtrip_registered/run.pyLocate this file in the curated companion packet; no public download is currently offered.
• runner · staged-not-public — synthetic-creativity-archive:2026-08-20_glp1_grant3_cross_field_invariants_registered/run.pyLocate this file in the curated companion packet; no public download is currently offered.

Author-declared entry points

• run.py in registered packets

Resolved code

Evidence record

Artifacts and frozen runs

Result artifacts

• RESULTS.md
• MANUAL_SEMANTIC_AUDIT.md
• auditable_proposal_clean.md

Run records