conexus 2.3.3__tar.gz → 2.3.4__tar.gz

{conexus-2.3.3 → conexus-2.3.4}/.claude-plugin/marketplace.json

@@ -9,7 +9,7 @@
       "name": "nx",
       "source": "./nx",
       "description": "Self-hosted three-tier knowledge management with 15 specialized agents, semantic search, and RDR decision tracking for Claude Code.",
-      "version": "2.3.3"
+      "version": "2.3.4"
     }
   ]
 }

{conexus-2.3.3 → conexus-2.3.4}/CHANGELOG.md

@@ -6,13 +6,22 @@ Versioning follows [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
 ## [Unreleased]
 
+## [2.3.4] - 2026-03-23
+
+### Fixed
+- **Python 3.14 compatibility** — `voyageai` uses Pydantic v1 compat layer which
+  is broken on Python ≥ 3.14. Moved `voyageai` from required to optional dependency
+  (`pip install conexus[cloud]`). Capped `requires-python` to `<3.14` so uv/pip
+  auto-select Python 3.13. All `import voyageai` sites guarded with clear error
+  messages pointing to `conexus[cloud]`.
+- **Unprefixed skill references in docs** — all `/rdr-create`, `/rdr-close`, etc.
+  corrected to `/nx:rdr-create`, `/nx:rdr-close` across 11 documentation files.
+
 ## [2.3.3] - 2026-03-23
 
 ### Fixed
-- **Python 3.14 compatibility** — `import voyageai` at module level in `t3.py`
-  crashed on Python 3.14 due to Pydantic v1 incompatibility in the voyageai SDK.
-  Import is now guarded with `try/except`; cloud mode raises a clear error, local
-  mode works without voyageai.
+- **Python 3.14 compatibility** (partial) — guarded `import voyageai` in `t3.py`
+  but missed `retry.py` and other import sites. Superseded by 2.3.4.
 
 ## [2.3.2] - 2026-03-22
 

{conexus-2.3.3 → conexus-2.3.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: conexus
-Version: 2.3.3
+Version: 2.3.4
 Summary: Self-hosted semantic search and knowledge management for LLM-driven development
 Project-URL: Homepage, https://github.com/Hellblazer/nexus
 Project-URL: Repository, https://github.com/Hellblazer/nexus
@@ -19,7 +19,7 @@ Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
 Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
 Classifier: Topic :: Software Development :: Libraries :: Application Frameworks
-Requires-Python: >=3.12
+Requires-Python: <3.14,>=3.12
 Requires-Dist: chromadb>=0.6
 Requires-Dist: click>=8.1
 Requires-Dist: docling>=2.76.0
@@ -33,7 +33,8 @@ Requires-Dist: pyyaml>=6.0
 Requires-Dist: structlog>=24.0
 Requires-Dist: tqdm>=4.65
 Requires-Dist: tree-sitter-language-pack==0.7.1
-Requires-Dist: voyageai>=0.2
+Provides-Extra: cloud
+Requires-Dist: voyageai>=0.2; extra == 'cloud'
 Provides-Extra: local
 Requires-Dist: fastembed>=0.7.0; extra == 'local'
 Description-Content-Type: text/markdown

{conexus-2.3.3 → conexus-2.3.4}/docs/historical/plans/2026-02-23-rdr-workflow-skills-design.md

@@ -17,7 +17,7 @@ Nexus provides the data infrastructure (T2 memory, T3 semantic search, agent orc
 ### Separation of Concerns
 
 - **`nx`** = data infrastructure (store, search, index, memory)
-- **`/rdr-*`** = workflow orchestration as Claude Code slash command skills (live in the `nx/` plugin)
+- **`/nx:rdr-*`** = workflow orchestration as Claude Code slash command skills (live in the `nx/` plugin)
 - **`bd`** = execution tracking (beads for implementation work)
 
 RDR skills invoke `nx` under the hood for persistence but present a workflow-oriented interface to the user. This keeps `nx` focused on data and avoids coupling planning methodology into the CLI.
@@ -58,19 +58,19 @@ Active RDRs live in the filesystem (for collaboration) with structured metadata
 
 **Why `docs__rdr__` not `knowledge__rdr__`:** The `docs__` namespace routes to `voyage-context-3` (Contextualized Contextual Embedding) which embeds each chunk with awareness of surrounding chunks — critical for RDRs where individual sections derive meaning from the broader document context. The `knowledge__` namespace uses the same model but `docs__` is semantically correct: RDRs are structured documents, not atomic knowledge entries.
 
-**T2 TTL policy:** All RDR records use `--ttl permanent`. RDRs can span months from creation to close. The default 30-day TTL would silently expire active RDR metadata, causing `/rdr-gate` to report zero assumptions and `/rdr-list` to omit active RDRs.
+**T2 TTL policy:** All RDR records use `--ttl permanent`. RDRs can span months from creation to close. The default 30-day TTL would silently expire active RDR metadata, causing `/nx:rdr-gate` to report zero assumptions and `/nx:rdr-list` to omit active RDRs.
 
 **T3 title uniqueness:** T3 document IDs are derived from `sha256(collection:title)`. The T3 title must include the project prefix (e.g., `NX-003 Semantic Search Pipeline`) to guarantee uniqueness across projects that share a `docs__rdr__` collection prefix.
 
 ### Lifecycle
 
 ```
-/rdr-create ──author──> /rdr-research ──refine──> /rdr-gate ──pass──> /rdr-close
+/nx:rdr-create ──author──> /nx:rdr-research ──refine──> /nx:rdr-gate ──pass──> /nx:rdr-close
                              ^                        |
                              └────fail, iterate────────┘
 
-/rdr-show    (inspect one)
-/rdr-list    (inspect all)
+/nx:rdr-show    (inspect one)
+/nx:rdr-list    (inspect all)
 ```
 
 Status transitions: `Draft → Final → Implemented | Reverted | Abandoned | Superseded`
@@ -104,9 +104,9 @@ Future enhancement: after `check_and_reindex()`, inspect `git diff --name-only`
 
 ## Skill Specifications
 
-### `/rdr-create`
+### `/nx:rdr-create`
 
-**Trigger:** User says "create an RDR", "new RDR", or `/rdr-create`
+**Trigger:** User says "create an RDR", "new RDR", or `/nx:rdr-create`
 
 **Agent:** None — mechanical scaffolding.
 
@@ -131,9 +131,9 @@ Future enhancement: after `check_and_reindex()`, inspect `git diff --name-only`
 
 ---
 
-### `/rdr-research`
+### `/nx:rdr-research`
 
-**Trigger:** User says "add research finding", "update RDR research", or `/rdr-research`
+**Trigger:** User says "add research finding", "update RDR research", or `/nx:rdr-research`
 
 **Agent:** `deep-research-synthesizer` for gathering evidence. `codebase-deep-analyzer` for code-specific questions.
 
@@ -165,9 +165,9 @@ Future enhancement: after `check_and_reindex()`, inspect `git diff --name-only`
 
 ---
 
-### `/rdr-gate`
+### `/nx:rdr-gate`
 
-**Trigger:** User says "gate this RDR", "finalization check", or `/rdr-gate`
+**Trigger:** User says "gate this RDR", "finalization check", or `/nx:rdr-gate`
 
 **Agent:** `substantive-critic` for Layer 3 AI critique.
 
@@ -209,9 +209,9 @@ Future enhancement: after `check_and_reindex()`, inspect `git diff --name-only`
 
 ---
 
-### `/rdr-close`
+### `/nx:rdr-close`
 
-**Trigger:** User says "close this RDR", "RDR done", or `/rdr-close`
+**Trigger:** User says "close this RDR", "RDR done", or `/nx:rdr-close`
 
 **Agents:**
 - `knowledge-tidier` (Haiku) — archives post-mortem to T3 (post-mortem only; bead creation done directly by skill)
@@ -239,7 +239,7 @@ Future enhancement: after `check_and_reindex()`, inspect `git diff --name-only`
 - T2 tracks an `archived: false` flag that can be retried independently
 - Each step emits clear status so a failure is diagnosable
 - The skill reports which steps completed and which failed
-- Failed steps can be retried by re-running `/rdr-close` (idempotent: checks T2 state before repeating completed steps)
+- Failed steps can be retried by re-running `/nx:rdr-close` (idempotent: checks T2 state before repeating completed steps)
 
 **If Reverted or Abandoned:**
 1. Prompt for reason (free text)
@@ -261,9 +261,9 @@ Future enhancement: after `check_and_reindex()`, inspect `git diff --name-only`
 
 ---
 
-### `/rdr-show`
+### `/nx:rdr-show`
 
-**Trigger:** User says "show RDR", or `/rdr-show`
+**Trigger:** User says "show RDR", or `/nx:rdr-show`
 
 **Agent:** None — read-only display.
 
@@ -279,9 +279,9 @@ Future enhancement: after `check_and_reindex()`, inspect `git diff --name-only`
 
 ---
 
-### `/rdr-list`
+### `/nx:rdr-list`
 
-**Trigger:** User says "list RDRs", or `/rdr-list`
+**Trigger:** User says "list RDRs", or `/nx:rdr-list`
 
 **Agent:** None — read-only query.
 
@@ -297,16 +297,16 @@ Future enhancement: after `check_and_reindex()`, inspect `git diff --name-only`
 
 | Skill | Agent | Role |
 |-------|-------|------|
-| `/rdr-create` | — | Mechanical scaffolding |
-| `/rdr-research add` | `deep-research-synthesizer` | Fan out across web, codebase, T3, DEVONthink to gather and classify evidence |
-| `/rdr-research add` (code) | `codebase-deep-analyzer` | Deep code exploration for code-specific research questions |
-| `/rdr-gate` (Layer 3) | `substantive-critic` | Structural flaws, contradictions, unvalidated assumptions, proportionality |
-| `/rdr-gate` (prior art) | `nx search --corpus docs__rdr` | Prefix fan-out across all repos' RDR collections |
-| `/rdr-close` (decompose) | — (direct) | Parse Implementation Plan into beads via `bd create` + `bd dep add` |
-| `/rdr-close` (archive) | `knowledge-tidier` | Archive post-mortem to `knowledge__rdr_postmortem__{repo}` |
-| `/rdr-close` (divergence) | `substantive-critic` | Classify divergence patterns if post-mortem created |
-| `/rdr-show` | — | Read-only display |
-| `/rdr-list` | — | Read-only query |
+| `/nx:rdr-create` | — | Mechanical scaffolding |
+| `/nx:rdr-research add` | `deep-research-synthesizer` | Fan out across web, codebase, T3, DEVONthink to gather and classify evidence |
+| `/nx:rdr-research add` (code) | `codebase-deep-analyzer` | Deep code exploration for code-specific research questions |
+| `/nx:rdr-gate` (Layer 3) | `substantive-critic` | Structural flaws, contradictions, unvalidated assumptions, proportionality |
+| `/nx:rdr-gate` (prior art) | `nx search --corpus docs__rdr` | Prefix fan-out across all repos' RDR collections |
+| `/nx:rdr-close` (decompose) | — (direct) | Parse Implementation Plan into beads via `bd create` + `bd dep add` |
+| `/nx:rdr-close` (archive) | `knowledge-tidier` | Archive post-mortem to `knowledge__rdr_postmortem__{repo}` |
+| `/nx:rdr-close` (divergence) | `substantive-critic` | Classify divergence patterns if post-mortem created |
+| `/nx:rdr-show` | — | Read-only display |
+| `/nx:rdr-list` | — | Read-only query |
 
 **Agent selection rationale:** The original design used `strategic-planner` (Opus) for bead decomposition. Rev 2 delegated to `knowledge-tidier` (Haiku). Rev 3 moved bead creation back into the skill itself — the Implementation Plan section has a defined structure (Phase N / Step N headings), the skill already has `Bash` access, and splitting bead creation from archival simplifies failure handling. `knowledge-tidier` is now only used for post-mortem archival to T3 (its core strength: knowledge organization).
 
@@ -326,7 +326,7 @@ docs/
         └── ...
 ```
 
-**Bootstrapping:** `/rdr-create` handles first-use initialization: creates `docs/rdr/` and `docs/rdr/post-mortem/` directories, copies templates, initializes the README index. No separate bootstrap command needed.
+**Bootstrapping:** `/nx:rdr-create` handles first-use initialization: creates `docs/rdr/` and `docs/rdr/post-mortem/` directories, copies templates, initializes the README index. No separate bootstrap command needed.
 
 **README index merge conflicts:** The auto-generated `docs/rdr/README.md` will conflict in multi-contributor workflows. This is acceptable — the file is trivially regenerable from T2 state and filesystem contents. Consider adding a `.gitattributes` entry marking it as auto-generated.
 
@@ -412,7 +412,7 @@ ttl_days: 0                                 # permanent
 **Cross-project search:** `resolve_corpus()` treats arguments containing `__` as exact collection names, not prefixes. Therefore `--corpus docs__rdr` does **not** fan out. The skills use collection enumeration for cross-project search:
 
 ```bash
-# In /rdr-gate Layer 3:
+# In /nx:rdr-gate Layer 3:
 collections=$(nx collection list | grep '^docs__rdr__')
 for col in $collections; do
   nx search "query" --corpus "$col" --n 5
@@ -421,7 +421,7 @@ done
 
 **Repo name length constraint:** ChromaDB enforces 63-character collection names. `docs__rdr__` is 11 characters, leaving 52 for the repo name. Repo names exceeding 52 characters are truncated, with the full name stored in document metadata.
 
-**Cold-start behavior:** On first use (no `docs__rdr__*` collections exist), `/rdr-gate` Layer 3 gracefully reports "No prior RDRs indexed. Cross-project prior-art search will improve as RDRs are indexed and closed." T3 search handles non-existent collections via `_ChromaNotFoundError` → `continue`.
+**Cold-start behavior:** On first use (no `docs__rdr__*` collections exist), `/nx:rdr-gate` Layer 3 gracefully reports "No prior RDRs indexed. Cross-project prior-art search will improve as RDRs are indexed and closed." T3 search handles non-existent collections via `_ChromaNotFoundError` → `continue`.
 
 ## Integration Points
 
@@ -429,17 +429,17 @@ done
 nx index rdr ──────────────────────> docs__rdr__{repo} (T3, semantic index)
                                      SemanticMarkdownChunker + voyage-context-3
 
-/rdr-create ─────────────────────> docs/rdr/NNN.md (filesystem)
+/nx:rdr-create ─────────────────────> docs/rdr/NNN.md (filesystem)
                                     nx memory put - ... --ttl permanent (T2)
 
-/rdr-research ──agent──> findings > docs/rdr/NNN.md (append to Research Findings)
+/nx:rdr-research ──agent──> findings > docs/rdr/NNN.md (append to Research Findings)
                                     nx memory put - ... --ttl permanent (T2, structured)
 
-/rdr-gate ──────agent──> critique > nx search --corpus docs__rdr (T3, prior art)
+/nx:rdr-gate ──────agent──> critique > nx search --corpus docs__rdr (T3, prior art)
              substantive-critic     nx memory put - ... (T2, status=Final)
                                     nx index rdr (T3, re-index updated content)
 
-/rdr-close ─────direct──> beads ──> bd create (epic + children)
+/nx:rdr-close ─────direct──> beads ──> bd create (epic + children)
                                     bd dep add (wiring)
                                     docs/rdr/post-mortem/NNN.md (if diverged)
            ─────direct──> state ──> nx memory put - ... (T2, status=closed)
@@ -456,22 +456,22 @@ For repos with existing RDRs (e.g., arcaneum with 19 RDRs) that predate this sys
 # Index existing RDRs for semantic search:
 nx index rdr /path/to/repo
 
-# Optionally seed T2 metadata for /rdr-list and /rdr-show:
-# A one-time /rdr-import skill could scan docs/rdr/*.md,
+# Optionally seed T2 metadata for /nx:rdr-list and /nx:rdr-show:
+# A one-time /nx:rdr-import skill could scan docs/rdr/*.md,
 # parse frontmatter, and create T2 records. Not in initial scope
 # but a natural follow-up.
 ```
 
-The `nx index rdr` command works immediately on any repo with `docs/rdr/*.md` — no migration needed for semantic search. T2 metadata is only needed for the structured query features (`/rdr-list --has-assumptions`, etc.) and can be populated incrementally.
+The `nx index rdr` command works immediately on any repo with `docs/rdr/*.md` — no migration needed for semantic search. T2 metadata is only needed for the structured query features (`/nx:rdr-list --has-assumptions`, etc.) and can be populated incrementally.
 
 ## What This Does NOT Cover
 
 - **RDR content authoring** — The skills manage lifecycle, not prose. The author writes the RDR.
-- **Automated research execution** — `/rdr-research add` records findings; the agent helps gather them but the human classifies.
+- **Automated research execution** — `/nx:rdr-research add` records findings; the agent helps gather them but the human classifies.
 - **PR integration** — RDR files are version-controlled and naturally participate in PRs, but there's no GitHub-specific automation.
 - **Cross-repo RDR dependencies** — T3 archive enables cross-project search, but there's no formal dependency graph between RDRs in different repos.
 - **Server auto-reindex of RDRs** — Deferred to v2. Currently requires manual `nx index rdr` after changes.
-- **Filesystem-T2 sync** — If a user edits an RDR's markdown metadata directly (e.g., changes status), T2 is not automatically updated. The skills are the intended interface; direct edits may cause T2/filesystem drift. A future `/rdr-sync` command could reconcile.
+- **Filesystem-T2 sync** — If a user edits an RDR's markdown metadata directly (e.g., changes status), T2 is not automatically updated. The skills are the intended interface; direct edits may cause T2/filesystem drift. A future `/nx:rdr-sync` command could reconcile.
 
 ## Revision History
 

{conexus-2.3.3 → conexus-2.3.4}/docs/historical/plans/2026-02-23-rdr-workflow-skills-impl-plan.md

@@ -2,7 +2,7 @@
 
 > **For Claude:** REQUIRED SUB-SKILL: Use superpowers:executing-plans to implement this plan task-by-task.
 
-**Goal:** Implement the RDR workflow automation system: `nx index rdr` CLI command, SessionStart hook, and 6 Claude Code slash command skills (`/rdr-create`, `/rdr-list`, `/rdr-show`, `/rdr-research`, `/rdr-gate`, `/rdr-close`).
+**Goal:** Implement the RDR workflow automation system: `nx index rdr` CLI command, SessionStart hook, and 6 Claude Code slash command skills (`/nx:rdr-create`, `/nx:rdr-list`, `/nx:rdr-show`, `/nx:rdr-research`, `/nx:rdr-gate`, `/nx:rdr-close`).
 
 **Architecture:** Python CLI command + hook script for infrastructure; markdown skill files for workflow orchestration. Skills invoke `nx` CLI and `bd` for state management. Storage: filesystem (docs/rdr/*.md) + T2 (structured metadata) + T3 (semantic index via `docs__rdr__{repo}`).
 
@@ -331,9 +331,9 @@ git commit -m "feat: bundle RDR and post-mortem templates as plugin resources"
 
 ---
 
-## Phase 4: Skills — Read-Only First (`/rdr-list`, `/rdr-show`)
+## Phase 4: Skills — Read-Only First (`/nx:rdr-list`, `/nx:rdr-show`)
 
-### Task 5: Create `/rdr-list` skill
+### Task 5: Create `/nx:rdr-list` skill
 
 **Files:**
 - Create: `nx/skills/rdr-list/SKILL.md`
@@ -345,7 +345,7 @@ git commit -m "feat: bundle RDR and post-mortem templates as plugin resources"
 name: rdr-list
 description: >
   List all RDRs in the current project with status, type, and priority.
-  Triggers: user says "list RDRs", "show all RDRs", or /rdr-list
+  Triggers: user says "list RDRs", "show all RDRs", or /nx:rdr-list
 allowed-tools: Read, Glob, Grep, Bash
 ---
 
@@ -354,7 +354,7 @@ allowed-tools: Read, Glob, Grep, Bash
 ## When This Skill Activates
 
 - User says "list RDRs", "show RDRs", "what RDRs exist"
-- User invokes `/rdr-list`
+- User invokes `/nx:rdr-list`
 - User asks about the state of planning documents
 
 ## Behavior
@@ -394,12 +394,12 @@ If the user specifies filters, apply them:
 ```bash
 cd /Users/hal.hildebrand/git/nexus
 git add nx/skills/rdr-list/SKILL.md
-git commit -m "feat: add /rdr-list skill for listing RDRs with status"
+git commit -m "feat: add /nx:rdr-list skill for listing RDRs with status"
 ```
 
 ---
 
-### Task 6: Create `/rdr-show` skill
+### Task 6: Create `/nx:rdr-show` skill
 
 **Files:**
 - Create: `nx/skills/rdr-show/SKILL.md`
@@ -411,7 +411,7 @@ git commit -m "feat: add /rdr-list skill for listing RDRs with status"
 name: rdr-show
 description: >
   Display detailed RDR information including content, status, research findings,
-  and linked beads. Triggers: user says "show RDR", "RDR details", or /rdr-show
+  and linked beads. Triggers: user says "show RDR", "RDR details", or /nx:rdr-show
 allowed-tools: Read, Glob, Grep, Bash
 ---
 
@@ -420,7 +420,7 @@ allowed-tools: Read, Glob, Grep, Bash
 ## When This Skill Activates
 
 - User says "show RDR 003", "RDR details", "what's in RDR 5"
-- User invokes `/rdr-show`
+- User invokes `/nx:rdr-show`
 - User asks about a specific RDR's content or status
 
 ## Behavior
@@ -456,7 +456,7 @@ allowed-tools: Read, Glob, Grep, Bash
 (not closed yet)
 ```
 
-6. If the RDR ID is not found, list available RDRs (delegate to `/rdr-list` behavior).
+6. If the RDR ID is not found, list available RDRs (delegate to `/nx:rdr-list` behavior).
 
 ## Notes
 
@@ -470,14 +470,14 @@ allowed-tools: Read, Glob, Grep, Bash
 ```bash
 cd /Users/hal.hildebrand/git/nexus
 git add nx/skills/rdr-show/SKILL.md
-git commit -m "feat: add /rdr-show skill for detailed RDR inspection"
+git commit -m "feat: add /nx:rdr-show skill for detailed RDR inspection"
 ```
 
 ---
 
-## Phase 5: Skills — Lifecycle (`/rdr-create`, `/rdr-research`)
+## Phase 5: Skills — Lifecycle (`/nx:rdr-create`, `/nx:rdr-research`)
 
-### Task 7: Create `/rdr-create` skill
+### Task 7: Create `/nx:rdr-create` skill
 
 **Files:**
 - Create: `nx/skills/rdr-create/SKILL.md`
@@ -489,7 +489,7 @@ git commit -m "feat: add /rdr-show skill for detailed RDR inspection"
 name: rdr-create
 description: >
   Scaffold a new RDR from template, assign sequential ID, register in T2, add to index.
-  Triggers: user says "create an RDR", "new RDR", or /rdr-create
+  Triggers: user says "create an RDR", "new RDR", or /nx:rdr-create
 allowed-tools: Read, Write, Edit, Glob, Grep, Bash
 ---
 
@@ -498,7 +498,7 @@ allowed-tools: Read, Write, Edit, Glob, Grep, Bash
 ## When This Skill Activates
 
 - User says "create an RDR", "new RDR", "start an RDR"
-- User invokes `/rdr-create`
+- User invokes `/nx:rdr-create`
 - User wants to document a technical decision before implementation
 
 ## Inputs
@@ -583,13 +583,13 @@ Created RDR PREFIX-NNN: "Title"
 File: docs/rdr/NNN-kebab-title.md
 Status: Draft
 
-Next: Fill in Problem Statement and Context, then use /rdr-research to add findings.
+Next: Fill in Problem Statement and Context, then use /nx:rdr-research to add findings.
 ```
 
 ## Does NOT
 
-- Create beads (that happens at `/rdr-close`)
-- Run validation (that happens at `/rdr-gate`)
+- Create beads (that happens at `/nx:rdr-close`)
+- Run validation (that happens at `/nx:rdr-gate`)
 - Commit (user decides when to commit)
 - Run `nx index rdr` (user can do this manually or it happens at gate/close)
 ```
@@ -599,12 +599,12 @@ Next: Fill in Problem Statement and Context, then use /rdr-research to add findi
 ```bash
 cd /Users/hal.hildebrand/git/nexus
 git add nx/skills/rdr-create/SKILL.md
-git commit -m "feat: add /rdr-create skill for scaffolding new RDRs"
+git commit -m "feat: add /nx:rdr-create skill for scaffolding new RDRs"
 ```
 
 ---
 
-### Task 8: Create `/rdr-research` skill
+### Task 8: Create `/nx:rdr-research` skill
 
 **Files:**
 - Create: `nx/skills/rdr-research/SKILL.md`
@@ -616,7 +616,7 @@ git commit -m "feat: add /rdr-create skill for scaffolding new RDRs"
 name: rdr-research
 description: >
   Add, track, and verify structured research findings for an RDR.
-  Triggers: user says "add research finding", "RDR research", or /rdr-research
+  Triggers: user says "add research finding", "RDR research", or /nx:rdr-research
 allowed-tools: Task, Read, Write, Edit, Glob, Grep, Bash
 ---
 
@@ -627,12 +627,12 @@ Optionally delegates to **deep-research-synthesizer** (sonnet) for evidence gath
 ## When This Skill Activates
 
 - User says "add research finding", "update RDR research", "verify assumption"
-- User invokes `/rdr-research`
+- User invokes `/nx:rdr-research`
 - User wants to record or classify a discovery during RDR planning
 
 ## Subcommands
 
-### `/rdr-research add <id>`
+### `/nx:rdr-research add <id>`
 
 **Inputs** (prompt the user):
 1. **Finding text**: What was discovered
@@ -667,7 +667,7 @@ Optionally delegates to **deep-research-synthesizer** (sonnet) for evidence gath
      *Source: source description*
    ```
 
-### `/rdr-research status <id>`
+### `/nx:rdr-research status <id>`
 
 1. List T2 entries: `nx memory list --project {repo}_rdr`
 2. Filter titles matching `NNN-research-*`
@@ -681,7 +681,7 @@ Optionally delegates to **deep-research-synthesizer** (sonnet) for evidence gath
      - [seq 6] "Latency under 100ms" (docs only)
    ```
 
-### `/rdr-research verify <id> <finding-seq>`
+### `/nx:rdr-research verify <id> <finding-seq>`
 
 1. Read T2 record: `nx memory get --project {repo}_rdr --title NNN-research-{seq}`
 2. Prompt for new classification (verified or documented) and updated verification method
@@ -707,14 +707,14 @@ When the user asks to *investigate* something (not just record a finding):
 ```bash
 cd /Users/hal.hildebrand/git/nexus
 git add nx/skills/rdr-research/SKILL.md
-git commit -m "feat: add /rdr-research skill for structured research tracking"
+git commit -m "feat: add /nx:rdr-research skill for structured research tracking"
 ```
 
 ---
 
-## Phase 6: Skills — Gate and Close (`/rdr-gate`, `/rdr-close`)
+## Phase 6: Skills — Gate and Close (`/nx:rdr-gate`, `/nx:rdr-close`)
 
-### Task 9: Create `/rdr-gate` skill
+### Task 9: Create `/nx:rdr-gate` skill
 
 **Files:**
 - Create: `nx/skills/rdr-gate/SKILL.md`
@@ -726,7 +726,7 @@ git commit -m "feat: add /rdr-research skill for structured research tracking"
 name: rdr-gate
 description: >
   Run the RDR finalization gate: structural validation, assumption audit, and AI critique.
-  Triggers: user says "gate this RDR", "finalization check", or /rdr-gate
+  Triggers: user says "gate this RDR", "finalization check", or /nx:rdr-gate
 allowed-tools: Task, Read, Write, Edit, Glob, Grep, Bash
 ---
 
@@ -737,7 +737,7 @@ Delegates Layer 3 to the **substantive-critic** agent (sonnet). See [registry.ya
 ## When This Skill Activates
 
 - User says "gate this RDR", "finalization check", "is this RDR ready?"
-- User invokes `/rdr-gate`
+- User invokes `/nx:rdr-gate`
 - User wants to validate an RDR before locking it as Final
 
 ## Input
@@ -857,12 +857,12 @@ Display the critique with specific sections to address. Status remains Draft.
 ```bash
 cd /Users/hal.hildebrand/git/nexus
 git add nx/skills/rdr-gate/SKILL.md
-git commit -m "feat: add /rdr-gate skill with 3-layer validation"
+git commit -m "feat: add /nx:rdr-gate skill with 3-layer validation"
 ```
 
 ---
 
-### Task 10: Create `/rdr-close` skill
+### Task 10: Create `/nx:rdr-close` skill
 
 **Files:**
 - Create: `nx/skills/rdr-close/SKILL.md`
@@ -874,7 +874,7 @@ git commit -m "feat: add /rdr-gate skill with 3-layer validation"
 name: rdr-close
 description: >
   Close an RDR: capture divergence, create post-mortem, decompose into beads, archive to T3.
-  Triggers: user says "close this RDR", "RDR done", or /rdr-close
+  Triggers: user says "close this RDR", "RDR done", or /nx:rdr-close
 allowed-tools: Task, Read, Write, Edit, Glob, Grep, Bash
 ---
 
@@ -885,7 +885,7 @@ Delegates decomposition and archival to the **knowledge-tidier** agent (haiku).
 ## When This Skill Activates
 
 - User says "close this RDR", "RDR done", "finish RDR"
-- User invokes `/rdr-close`
+- User invokes `/nx:rdr-close`
 - Implementation is complete and the RDR should be finalized
 
 ## Inputs
@@ -968,7 +968,7 @@ Display the bead tree for user confirmation before proceeding.
         close_reason: "Implemented", epic_bead: "{bead_id}", archived: true)
    EOF
    ```
-   If T3 archive fails, set `archived: false` — retryable by re-running `/rdr-close`
+   If T3 archive fails, set `archived: false` — retryable by re-running `/nx:rdr-close`
 
 2. Update status in RDR markdown metadata
 3. Regenerate `docs/rdr/README.md` index
@@ -1015,7 +1015,7 @@ nx store put docs/rdr/post-mortem/NNN-*.md --collection docs__rdr__{repo} \
 The close operation performs multiple state mutations. If any step fails:
 - Each step emits clear status (e.g., "T2 updated ✓", "Beads created ✓", "T3 archive ✗ FAILED")
 - T2 `archived` flag tracks whether T3 archival succeeded
-- Re-running `/rdr-close` is idempotent: checks T2 state and skips completed steps
+- Re-running `/nx:rdr-close` is idempotent: checks T2 state and skips completed steps
 - If bead creation partially fails, report which beads were created and which failed
 
 ## Does NOT
@@ -1030,7 +1030,7 @@ The close operation performs multiple state mutations. If any step fails:
 ```bash
 cd /Users/hal.hildebrand/git/nexus
 git add nx/skills/rdr-close/SKILL.md
-git commit -m "feat: add /rdr-close skill with post-mortem, bead decomposition, T3 archive"
+git commit -m "feat: add /nx:rdr-close skill with post-mortem, bead decomposition, T3 archive"
 ```
 
 ---
@@ -1053,7 +1053,7 @@ No new agents are needed — the RDR skills dispatch to existing agents (substan
 
 rdr_skills:
   rdr-create:
-    slash_command: /rdr-create
+    slash_command: /nx:rdr-create
     description: "Scaffold new RDR from template, assign ID, register in T2"
     triggers:
       - "create an RDR"
@@ -1061,7 +1061,7 @@ rdr_skills:
       - "start planning document"
 
   rdr-research:
-    slash_command: /rdr-research
+    slash_command: /nx:rdr-research
     description: "Add, track, and verify structured research findings"
     triggers:
       - "add research finding"
@@ -1070,7 +1070,7 @@ rdr_skills:
     dispatches_to: [deep-research-synthesizer, codebase-deep-analyzer]
 
   rdr-gate:
-    slash_command: /rdr-gate
+    slash_command: /nx:rdr-gate
     description: "Run finalization gate: structural + assumption + AI critique"
     triggers:
       - "gate this RDR"
@@ -1079,7 +1079,7 @@ rdr_skills:
     dispatches_to: [substantive-critic]
 
   rdr-close:
-    slash_command: /rdr-close
+    slash_command: /nx:rdr-close
     description: "Close RDR with post-mortem, bead decomposition, T3 archive"
     triggers:
       - "close this RDR"
@@ -1088,14 +1088,14 @@ rdr_skills:
     dispatches_to: [knowledge-tidier]
 
   rdr-list:
-    slash_command: /rdr-list
+    slash_command: /nx:rdr-list
     description: "List all RDRs with status, type, priority"
     triggers:
       - "list RDRs"
       - "show all RDRs"
 
   rdr-show:
-    slash_command: /rdr-show
+    slash_command: /nx:rdr-show
     description: "Display detailed RDR information"
     triggers:
       - "show RDR"
@@ -1208,10 +1208,10 @@ If any fixes were needed during smoke testing, commit them.
 1. `feat: add nx index rdr command for RDR document indexing`
 2. `feat: add SessionStart hook for RDR auto-detection`
 3. `feat: bundle RDR and post-mortem templates as plugin resources`
-4. `feat: add /rdr-list skill for listing RDRs with status`
-5. `feat: add /rdr-show skill for detailed RDR inspection`
-6. `feat: add /rdr-create skill for scaffolding new RDRs`
-7. `feat: add /rdr-research skill for structured research tracking`
-8. `feat: add /rdr-gate skill with 3-layer validation`
-9. `feat: add /rdr-close skill with post-mortem, bead decomposition, T3 archive`
+4. `feat: add /nx:rdr-list skill for listing RDRs with status`
+5. `feat: add /nx:rdr-show skill for detailed RDR inspection`
+6. `feat: add /nx:rdr-create skill for scaffolding new RDRs`
+7. `feat: add /nx:rdr-research skill for structured research tracking`
+8. `feat: add /nx:rdr-gate skill with 3-layer validation`
+9. `feat: add /nx:rdr-close skill with post-mortem, bead decomposition, T3 archive`
 10. `feat: register RDR skills in registry, add RDR awareness to analyzer`

{conexus-2.3.3 → conexus-2.3.4}/docs/memory-and-tasks.md

@@ -40,7 +40,7 @@ The plugin wires beads into the session lifecycle:
 - **SessionStart** also shows ready beads (unblocked work) via `bd ready`
 - **SubagentStart** injects the active bead so spawned agents know what
   task they're continuing
-- **RDR close** (`/rdr-close`) decomposes a decision into beads — one epic
+- **RDR close** (`/nx:rdr-close`) decomposes a decision into beads — one epic
   for the overall effort, plus task beads for each implementation step
 - **Branch naming** ties git branches to beads: `feature/<bead-id>-<description>`
 
@@ -55,7 +55,7 @@ the other.
 
 When you use both, the connections are automated:
 
-- `/rdr-close` creates beads (epic + task beads) for implementation tracking.
+- `/nx:rdr-close` creates beads (epic + task beads) for implementation tracking.
   The `epic_bead` field in each RDR's T2 metadata provides a machine-readable
   link from decision to work items.
 - RDR decisions surface as prior art during planning via `nx search "topic"` against the knowledge corpus.