Overview
Agentic software can preserve task state while still forgetting the consequences of prior actions. Codomyrmex studies a narrow control-plane question: after a caller reports a failed action at one software location, can the system deterministically increase friction for a materially similar proposal at that location without changing an unrelated target? Its Colony Control Plane records consequence reports and couples them to target-indexed signal pressure, agent trust, role labels, resource accounting, adversarial checks, and an explicit EXECUTE/HOLD/REFUSE gate. The implementation comprises 8 cooperating subsystems. The ordinary Model Context Protocol path remains caller-reported and unattested. Optional and required `ColonyKernel` attestation modes instead bind proposal, verdict, authorization, execution receipt, and outcome in a signed, hash-linked local ledger. That ledger protects lifecycle linkage but does not independently observe external actuation or establish deployment safety. Consequence records can use file-backed SQLite; the default MCP kernel and signal field remain process-local. Evaluation is limited to implementation properties and controlled fixtures. At composition time, the scoped Colony Kernel surface contains 819 passing tests with 76.6% branch coverage, 0 Ruff errors, and 0 ty diagnostics. A paired deterministic replay moves the same-target proposal from 0.875/EXECUTE to 0.725/HOLD after a reported failure while leaving an unrelated target unchanged. Separate fixtures exercise trust promotion, bounded arithmetic, linear signal decay, local attestation integrity, and interface behavior. These results support reproducible software contracts, not ecological optimality, calibrated risk, production harm reduction, or generalization to external workloads. The report contributes the typed control plane, transparent gate, coupled local feedback, authenticated local lifecycle option, and source-bound publication workflow. Generated variables, figures, citations, claim boundaries, and release receipts tie the rendered report to the evaluated checkout. End-to-end external-actuation attestation, restart-persistent field storage, representative benchmarks, and independent deployment validation remain open.
---
Associated artifacts
GitHub release: v1.3.0-paper (https://github.com/docxology/codomyrmex/releases/tag/v1.3.0-paper)
DOI: https://doi.org/10.5281/zenodo.21750801
Zenodo: https://zenodo.org/records/21750801
PDF SHA-256: eda76ad12a50bce01b113894c785e0915b6ba367f5bf67d17c8f586416102b93\nConcept DOI: https://doi.org/10.5281/zenodo.21750800\nVersion DOI: https://doi.org/10.5281/zenodo.21750801
Overview source: Curated paper metadata.
Methods and contributions
Read the source for the full argument, qualifications, and evidence.
Findings and contributions
- The scoped Colony Kernel surface has 819 passing tests with 76.6% branch coverage, 0 Ruff errors, and 0 ty diagnostics.
- After a reported failure, the paired replay moves the same-target proposal from 0.875/EXECUTE to 0.725/HOLD while an unrelated target is unchanged.
- The author states these results support reproducible software contracts, not ecological optimality, calibrated risk, production harm reduction, or generalization to external workloads.
- The proposed comparative benchmark has not been run, and no raw trial traces are included in this release.
- Default state is in memory, so the artifact does not support claims that pressure, trust, or gate state survive process restarts, model swaps, or deployment across machines.
Methods
- Colony Control Plane with 8 named subsystems — The implementation separates signal storage, resource accounting, actuation gating, consequence records, role inference, pruning, deterministic falsification, and integration.
- Ternary EXECUTE/HOLD/REFUSE gate with weighted bounded score — Budget, effective local hazard, trust credit, and proposal completeness are combined into a weighted score with hard overrides that route each proposal.
- Stigmergic signal field with FAILURE and RISK pressure at target locations — Reported failures and prospective risks are stored separately in a process-local field, and the gate uses their maximum as the local hazard.
- Contract test suite using real subsystem instances — Tests check same-target inhibition, cross-target isolation, linear decay recovery, score bounds, trust updates, and interface behavior.
- Paired deterministic replay of same-target vs. unrelated-target proposals — Identical proposals are evaluated with and without a reported failure at the target location to test failure-to-gate coupling.
Summary sources: Paper metadata and evidence · Extracted source text.
Document downloads
PDFs are hosted here; supporting sources may use pinned archives.
- codomyrmex-1.3.0-content.pdf PDF · 6.56 MiB
- codomyrmex-1.3.0.pdf PDF · 6.56 MiB · extracted-text source
These files may be different versions or companion documents. The archive does not identify a latest edition.
Citation
Citation metadata follows the unified bibliography.
Catalog details and resources
Related works
Ranked by recorded methods, topics, contributors and associated software, with contributor roles and domain-only matches labelled. Metadata overlap is not a citation or a scientific equivalence.