ORG_DISTRIBUTION.md

Organization distribution

This repository is the canonical organization-wide skill library for The Interdependency.

Canonical source

All organization-level agent skill edits should happen here first.

Repo-local copies may exist under:

.agents/skills/

Propagation PRs should cite this repository and the source commit SHA.

Installed skills

• msdmd/ — Module Self-Declared Metadata Markdown
• doc-build/ — documentation coverage metadata blocks
• cap-build/ — capability inventory metadata blocks
• deps-build/ — dependency topology metadata blocks
• owner-build/ — module stewardship metadata blocks
• test-build/ — contract-test metadata blocks
• meta-module-build/ — metadata-first module scaffolding
• risk-boundary-build/ — runtime risk and permission boundary metadata blocks
• ratios/ — module composition ratio verification
• canon/ — canonical-source and doctrine maintenance
• domain-claims/ — domain-first lexical standing, collision checks, and semantic provenance gating
• visitor-intro/ — onboarding tour for newcomers landing at any org repo
• char-compress/ — skill-lib-owned bone/flesh context compression for handoffs and skill writing; historical notation is not current UCNS mathematics
• agent-instantiation/ — a0/a0ucns agent spawn/fork/merge lifecycle methodology
• a0p-instancing/ — a0-betatest (a0p) per-user CRUD + native-ZFAE instancing methodology
• manifest/ — living-spec generation
• llms-build/ — root llms.txt generation from LLMS blocks
• typed-meta-frontend/ — TypeScript self-building frontend generation from backend module metadata
• plain-lens/ — plain-language, multi-lens companion views of dense canonical text
• thought-lens/ — raw-thought to audience-legible translation with claim-kernel fidelity and back-translation checks
• meta/ — consultation router for current METAPAT authority; no frozen doctrine copy
• gonol-build/ — UCNS gonol objects/constructors/geometry + Stack language-construction research, closure, atomic participation, replay, and honest continuation boundaries; EDCM is measurement/evaluation only
• ucns-option-selection/ — fail-closed scoped UCNS option comparison, selection, ratification, non-transfer, rollback, and decision receipts
• epac-selection-display/ — exact provisional EPAC target and representation selection with receipt-backed display, status preservation, and a read-only WebMCP handoff boundary
• the-interdependency/ — org-wide workflow protocol and usage-guidance doctrine for The Interdependency projects
• interdependent-work-graph/ — cross-repository identity, authority, coordination, and shared stack-manifest doctrine
• stack-update/ — fail-closed structural stack update protocol; keeps authority, relation, lifecycle, provenance, manifests, BASE records, and work-graph identity coherent in one transaction
• project-incubation-graduation/ — incubation, qualification, extraction, release, reconsumption, and implementation-authority graduation doctrine
• distributed-publication/ — provenance-bearing materialization of one ordered publication from independently owned source units
• loop-eng/ — closed-loop engineering doctrine for repeatable Discover→Plan→Execute→Verify→Iterate workflows
• fresh-making/ — deterministic derivation freshness, minimal affected rebuild closure, executor-independent restoration, verification, and receipts
• action-calibration/ — action sizing doctrine for minimal decisive experiments, maximal coherent programs, prerequisite repair, and immediate containment
• repo-audit-repair/ — evidence-led repository audit, classified findings, authorized repair, and terminal verification
• skill-build/ — skill authoring, compliance, and individualized test-suite question workflow
• skill-usage/ — evidence-bearing local invocation counts and maturity designations
• ssh-automation/ — fail-closed SSH scripting and copy-paste automation doctrine
• vm-mcp/ — private VM MCP control plane with loopback-only runtime, non-root shell, systemd host-write confinement, and private-tunnel deployment
• sql-queries/ — warehouse SQL authoring doctrine (imported, Apache-2.0 — see ATTRIBUTION.md)
• statistical-analysis/ — statistical methods doctrine (imported, Apache-2.0 — see ATTRIBUTION.md)
• explore-data/ — dataset profiling doctrine (imported, Apache-2.0 — see ATTRIBUTION.md)
• validate-data/ — analysis QA doctrine (imported, Apache-2.0 — see ATTRIBUTION.md)
• data-visualization/ — chart-building doctrine (imported, Apache-2.0 — see ATTRIBUTION.md)

Target repos

Active vendoring consumers — carry a top-level .agents/skills/<skill>/ subset copied from here. These are exactly the repos the scheduled drift detector (.github/workflows/consumer-drift.yml) checks:

• The-Interdependency/a0
• The-Interdependency/ucns
• The-Interdependency/edcm
• The-Interdependency/interdependent-lib
• The-Interdependency/aimmh
• The-Interdependency/ai-tiw
• The-Interdependency/eml_ucns
• The-Interdependency/zfae
• The-Interdependency/pcea
• The-Interdependency/a0-betatest
• The-Interdependency/metapat
• The-Interdependency/ptcna
• The-Interdependency/pubskill-lib
• The-Interdependency/epac
• The-Interdependency/stack

Targets not in the drift matrix (do not vendor a top-level subset yet, so --require-vendored would fail them):

• The-Interdependency/a0ucns — an aggregator that embeds whole copies of other repos rather than vendoring a top-level .agents/skills/ subset. Its nested embeds carry their own copies; re-sync those from their source repos.

Archived or superseded — not active drift consumers:

• The-Interdependency/edcmbone — archived; maintained EDCM work lives in edcm
• The-Interdependency/PTCA, The-Interdependency/pcna (→ ptcna)

Add a repo to the active list — and the drift matrix — once it carries a canonical .agents/skills/ subset.

Collection points

Every consuming repo should eventually carry a root collection point:

<reponame>_msdmd.ts

Use python -m msdmd.collect --root . --repo <repo> --out <reponame>_msdmd.ts when the repo can run the collector locally. A provisional hand-seeded collection point is allowed only when it records a hmmm gap explaining what local generation still needs.

skill-lib_msdmd.ts is the root collection point for this canonical repo.

Propagation checklist

Use docs/propagation-checklist.md for the concrete source-change → target-repo PR sequence. Use docs/runner-config-guidance.md before judging large or artifact-heavy repos; frozen research artifacts, archives, generated trees, and vendored .agents/skills/ copies should not pollute the denominator.

Rule

Before assigning a stack-level task to one repository, agents should read:

.agents/skills/interdependent-work-graph/SKILL.md

Resolve the exact participating repository and evidence-source identities first. Repository boundaries remain authority and provenance boundaries, not agent-attention boundaries.

Before changing the structure of The-Interdependency/stack — including participant, pin, authority, relation, workspace, BASE, extraction/graduation, or architecture projections — agents should read:

.agents/skills/stack-update/SKILL.md

Load interdependent-work-graph with it. Treat the mutation as one coherent transaction: update every affected authority/provenance projection, remove superseded claims, recompute the work-graph digest, and require the stack consistency checker to pass before merge.

Before deciding whether a component born inside a stack, integration, laboratory, or incubator repository should become an independent repository/package, or before extracting, publishing, reconsuming, or declaring such a component graduated, agents should read:

.agents/skills/project-incubation-graduation/SKILL.md

Load interdependent-work-graph once the transition crosses repository boundaries. Graduation transfers implementation/public-contract authority only after qualification, release, and downstream reconsumption; it does not transfer semantic, proof, theorem, measurement, certification, or empirical status.

Before assembling one textbook, report, standard, corpus, archive, or public reading sequence from source-owned units distributed across repositories or independently owned files, agents should read:

.agents/skills/distributed-publication/SKILL.md

Load interdependent-work-graph with it. Preserve ordered source identity, source-local licenses and statuses, correction routing, fail-closed production retrieval, explicit fallback, static reading access, and provenance in the published build artifact.

Before creating a new module, route, service, adapter, schema, worker, engine, UI panel, migration, or experiment, agents should read:

.agents/skills/meta-module-build/SKILL.md

New module work should start with a MODULE_BUILD block. Unknown fields must be marked hmmm, not guessed.

Before creating or maintaining a root llms.txt, agents should read:

.agents/skills/llms-build/SKILL.md

Root LLM instructions should be declared in source LLMS blocks and generated with the llms-build runner, not hand-maintained as separate doctrine.

Before translating raw, recursive, context-heavy, fragmentary, coined, or private-language thought for another human audience or public surface, agents should read:

.agents/skills/thought-lens/SKILL.md

Recover and freeze the claim kernel before changing vocabulary. Reduce the context required from the reader without strengthening, flattening, or silently completing the thought; use hmmm where a bridge remains unresolved. Use plain-lens instead when the source is already a stable dense document.

Before promoting a word or phrase into canon, a theorem, schema, ontology, encoding, or other semantic control surface, agents should read:

.agents/skills/domain-claims/SKILL.md

Establish the applicable domain-qualified sense and resolve collisions before attaching provenance or authorizing structure.

Before constructing, reviewing, replaying, or extending UCNS gonols, agents should read:

.agents/skills/gonol-build/SKILL.md

Resolve the current UCNS source and evidence identities first. Preserve closure and atomic promotion, require explicit occurrence-addressed function plans, and do not restore the superseded gonal-morphology language model.

Before selecting among UCNS options or declaring a scoped winner, agents should read:

.agents/skills/ucns-option-selection/SKILL.md

Freeze scope, candidates, hard gates, evidence, policies, authority, and ratification before outcome comparison. Do not let scores compensate for failed gates or transfer a scoped result into universal UCNS canon.

Before selecting an EPAC artifact and representation for a human-facing or WebMCP display, agents should read:

.agents/skills/epac-selection-display/SKILL.md

Pin the provisional EPAC source, exact target, receipt, and available renderer; preserve research standings, nonclaims, sealed-comparison boundaries, and hmmm; and keep the WebMCP server a read-only registry/handoff rather than an EPAC executor.

Before choosing between the smallest decisive action and a maximal coherent program, selecting the highest-leverage next step under constrained time, attention, money, compute, or coordination, or deciding whether a bounded falsifier should precede a full build, agents should read:

.agents/skills/action-calibration/SKILL.md

Name the decision, preserve load-bearing invariants, compare complete cost vectors, freeze outcome-conditioned escalation rules, and let loop-eng execute the selected bounded loop.

Before auditing or auditing and repairing an existing repository, agents should read:

.agents/skills/repo-audit-repair/SKILL.md

Resolve exact repository identity and authority first; select checks from actual repository claims; classify defects separately from environment, external service, policy, and hmmm; repair only the owning layer under the user's authorization; and verify merge, release, and deployment as distinct terminal states when applicable.

Before writing, reviewing, or troubleshooting repeatable SSH automation or a large terminal paste that contains SSH, agents should read:

.agents/skills/ssh-automation/SKILL.md

Fail closed on endpoint identity and host-key trust, preserve local/remote shell boundaries, and keep bulk pasted error handling inside a child shell.

Before giving an MCP-capable agent operational contact with a private VM, agents should read:

.agents/skills/vm-mcp/SKILL.md

Keep SSH/OS Login credentials outside the model path, install the service loopback-only and non-root, establish private authenticated transport, prove read-only contact before enabling shell execution, and add privileged administration only as named bounded capabilities rather than a generic root shell.

Existing files are not retroactively noncompliant merely because they predate this skill.

hmmm

Target-repo propagation is not complete just because skill-lib is updated. Each target repo still needs a repo-local propagation PR, source commit SHA, collection point, and local verification record.