@moreih29/nexus-core 0.15.1 → 0.16.0

package/dist/hooks/agent-bootstrap.js

@@ -100,6 +100,22 @@ ${ruleContent}
 `) };
 };
 var handler_default = handler;
-export {
-  handler_default as default
-};
+
+// ../../../../../tmp/nexus-hook-entry-agent-bootstrap-1776671215246/agent-bootstrap-entry.ts
+import { readFileSync as readFileSync2 } from "node:fs";
+async function main() {
+  let raw = "";
+  try {
+    raw = readFileSync2(0, "utf-8");
+  } catch {}
+  const input = raw ? JSON.parse(raw) : {};
+  const result = await handler_default(input);
+  if (result != null && result !== undefined) {
+    process.stdout.write(JSON.stringify(result));
+  }
+}
+main().then(() => process.exit(0), (err) => {
+  process.stderr.write(String(err?.stack ?? err) + `
+`);
+  process.exit(1);
+});

package/dist/hooks/agent-finalize.js

@@ -159,6 +159,22 @@ Subagent "${agent_type}" finished. Tasks still pending with this role: ${ids}. R
   }
 };
 var handler_default = handler;
-export {
-  handler_default as default
-};
+
+// ../../../../../tmp/nexus-hook-entry-agent-finalize-1776671215240/agent-finalize-entry.ts
+import { readFileSync as readFileSync2 } from "node:fs";
+async function main() {
+  let raw = "";
+  try {
+    raw = readFileSync2(0, "utf-8");
+  } catch {}
+  const input = raw ? JSON.parse(raw) : {};
+  const result = await handler_default(input);
+  if (result != null && result !== undefined) {
+    process.stdout.write(JSON.stringify(result));
+  }
+}
+main().then(() => process.exit(0), (err) => {
+  process.stderr.write(String(err?.stack ?? err) + `
+`);
+  process.exit(1);
+});

package/dist/hooks/prompt-router.js

@@ -7295,6 +7295,22 @@ var handler = async (input) => {
 `) };
 };
 var handler_default = handler;
-export {
-  handler_default as default
-};
+
+// ../../../../../tmp/nexus-hook-entry-prompt-router-1776671215226/prompt-router-entry.ts
+import { readFileSync as readFileSync2 } from "node:fs";
+async function main() {
+  let raw = "";
+  try {
+    raw = readFileSync2(0, "utf-8");
+  } catch {}
+  const input = raw ? JSON.parse(raw) : {};
+  const result = await handler_default(input);
+  if (result != null && result !== undefined) {
+    process.stdout.write(JSON.stringify(result));
+  }
+}
+main().then(() => process.exit(0), (err) => {
+  process.stderr.write(String(err?.stack ?? err) + `
+`);
+  process.exit(1);
+});

package/dist/hooks/session-init.js

@@ -16,6 +16,22 @@ var handler = async (input) => {
   writeFileSync(join(sessionDir, "tool-log.jsonl"), "");
 };
 var handler_default = handler;
-export {
-  handler_default as default
-};
+
+// ../../../../../tmp/nexus-hook-entry-session-init-1776671215218/session-init-entry.ts
+import { readFileSync } from "node:fs";
+async function main() {
+  let raw = "";
+  try {
+    raw = readFileSync(0, "utf-8");
+  } catch {}
+  const input = raw ? JSON.parse(raw) : {};
+  const result = await handler_default(input);
+  if (result != null && result !== undefined) {
+    process.stdout.write(JSON.stringify(result));
+  }
+}
+main().then(() => process.exit(0), (err) => {
+  process.stderr.write(String(err?.stack ?? err) + `
+`);
+  process.exit(1);
+});

package/dist/manifests/opencode-manifest.json

@@ -6,7 +6,7 @@
         "SessionStart"
       ],
       "matcher": "*",
-      "handlerPath": "../assets/hooks/session-init/handler.js",
+      "handlerPath": "../hooks/session-init.js",
       "priority": 0,
       "timeout": 10
     },
@@ -16,7 +16,7 @@
         "UserPromptSubmit"
       ],
       "matcher": "*",
-      "handlerPath": "../assets/hooks/prompt-router/handler.js",
+      "handlerPath": "../hooks/prompt-router.js",
       "priority": 0,
       "timeout": 10
     },
@@ -26,7 +26,7 @@
         "SubagentStop"
       ],
       "matcher": "*",
-      "handlerPath": "../assets/hooks/agent-finalize/handler.js",
+      "handlerPath": "../hooks/agent-finalize.js",
       "priority": 0,
       "timeout": 10
     },
@@ -36,7 +36,7 @@
         "SubagentStart"
       ],
       "matcher": "*",
-      "handlerPath": "../assets/hooks/agent-bootstrap/handler.js",
+      "handlerPath": "../hooks/agent-bootstrap.js",
       "priority": 0,
       "timeout": 10
     }

package/dist/opencode/.opencode/skills/nx-init/SKILL.md

@@ -0,0 +1,189 @@
+---
+description: "Project onboarding — scan, mission, essentials, context generation"
+---
+## Role
+
+Scans the project and builds Nexus knowledge in the flat .nexus/ structure. On first run, performs a 5-step full onboarding sequence.
+
+## Constraints
+
+- NEVER modify source code. Slimming down the instruction file beyond the project section is not this skill's responsibility.
+- NEVER infer or guess information that cannot be confirmed from code — do not write it to context/.
+- NEVER store secrets (API keys, credentials, etc.) in knowledge files.
+- NEVER overwrite existing files without `--reset`. On resume, preserve existing files intact.
+- Project section in the instruction file MUST go through user confirmation before writing.
+- NEVER reference or create identity/, codebase/, reference/, or core/ paths.
+- Essentials section MUST NOT exceed 10 lines. If more items are needed, move lower-priority ones to .nexus/context/.
+
+## Guidelines
+
+## Trigger
+
+- Manual trigger — full onboarding (or resume). See harness docs: slash_command_display.
+- Manual trigger with `--reset` flag — back up existing `.nexus/` knowledge and re-onboard. See harness docs: slash_command_display.
+- Manual trigger with `--reset --cleanup` flags — show backup list + selective deletion. See harness docs: slash_command_display.
+
+---
+
+## Modes
+
+### First Run (no `.nexus/` flat structure)
+
+Automatically runs the 5-step full onboarding.
+
+Detection: `.nexus/context/`, `.nexus/memory/`, `.nexus/state/`, `.nexus/rules/` do not exist.
+
+### Resume (`.nexus/` partially exists)
+
+Check existing state and resume from the first incomplete step.
+
+### Reset (`--reset`)
+
+Back up existing `.nexus/` knowledge directories to `.nexus/bak.{timestamp}/`, then enter First Run.
+
+### Cleanup (`--reset --cleanup`)
+
+Show backup directory list, let user select backups to delete.
+
+---
+
+## Process
+
+### Phase 0: Mode Detection
+
+```
+IF --reset --cleanup flag:
+  Show list of .nexus/bak.*/ directories
+  Prompt user with options via `question({ question: "Select a backup to delete (or cancel)", choices: [<backup list...>, {label: Cancel, description: "Exit without changes"}] })`.
+  Delete selected backup and exit
+
+ELSE IF --reset flag:
+  Move .nexus/{memory,context,state,rules}/ → .nexus/bak.{timestamp}/
+  Inform: "Existing knowledge has been backed up to .nexus/bak.{timestamp}/. Starting re-onboarding."
+  → Enter First Run
+
+ELSE IF .nexus/context/ exists:
+  → Enter Resume (check existing steps and resume)
+
+ELSE:
+  → Enter First Run (from Step 1)
+```
+
+---
+
+## Steps
+
+### Step 1: Project Scan
+
+Auto-detect code structure and tech stack. Create the flat `.nexus/` directory structure if it does not exist.
+
+Create directories (using shell command execution):
+- `.nexus/memory/`
+- `.nexus/context/`
+- `.nexus/state/`
+- `.nexus/rules/`
+
+Collected items:
+- **Directory structure**: top-level layout, major modules/packages
+- **Tech stack**: language, framework, runtime (package.json, Cargo.toml, pyproject.toml, go.mod, build.gradle, etc.)
+- **Build/test system**: scripts, CI configuration
+- **Existing docs**: CLAUDE.md, README.md, docs/, .cursorrules, etc.
+- **git context**: recent commits, branch structure, contributors
+
+Output: scan summary (language, framework, structure overview)
+
+For large projects (10+ top-level directories or 100+ files), consider spawning an Explore subagent for parallel scanning to reduce Lead context usage.
+
+### Step 2: Mission + Essentials (Interactive)
+
+Using the Step 1 scan results, draft a Mission statement (1–2 lines) and an Essentials list, then present both to the user for confirmation in a single pass.
+
+#### Essentials Guidelines
+
+Essentials are agent-critical facts — things an agent would get wrong if it didn't know them. Apply this judgment criterion: **"Would an agent produce wrong results without knowing this?"** Yes → Essentials. No → .nexus/context/.
+
+Draft from these five categories (include only what applies to this project):
+
+- **Tech stack** — runtime, language, package manager, core framework. Flag non-default tools (e.g. bun instead of npm, deno instead of node).
+- **Workflow** — build, test, deploy commands. Must-follow procedures such as required lint or type-check steps before committing.
+- **Constraints** — forbidden tools, patterns, or approaches. Directories or files that must not be modified.
+- **Domain** — target audience, required terminology or tone, compliance or regulatory constraints, methodology for research projects.
+- **Conventions** — naming, structure, or style that deviate from general defaults. Project-specific patterns an agent would not infer.
+
+Do not include items that are standard defaults for the detected tech stack. Do not exceed 10 lines total in the Essentials section.
+
+#### Draft Presentation
+
+Present the full draft to the user in this format:
+
+```
+The following will be added to the instruction file (see harness docs: instruction_file) (existing content will not be changed):
+
+<!-- PROJECT:START -->
+## {project-name}
+
+{mission 1-2 lines}
+
+### Essentials
+- {auto-detected item}
+- {auto-detected item}
+<!-- PROJECT:END -->
+
+Any changes?
+```
+
+Wait for the user to confirm or provide edits. Apply all changes in one pass — do not ask about Mission and Essentials separately.
+
+After confirmation, write the section into the instruction file inside markers using the harness's file-editing primitive. If the instruction file already contains `<!-- PROJECT:START -->` markers, replace the content between them. If the instruction file does not exist, create it with the markers.
+
+### Step 3: Context Knowledge Auto-Generation
+
+Analyze Step 1 scan results to generate context knowledge documents in `.nexus/context/`.
+
+Principles:
+- File names and content are decided freely based on project characteristics. No fixed templates.
+- Existing docs are information sources only — do not replicate their structure verbatim.
+- Do not guess content that cannot be confirmed from code.
+- Typically 1-3 files are sufficient. More files are not better.
+- **Generate abstract-level content only** — design patterns, architecture direction, module relationships, conventions. Do NOT include code-level details such as file listings, function signatures, or import maps. Those can be read directly from code.
+
+Generation targets (select and name based on what the project actually needs):
+- Development stack (languages, frameworks, runtimes, key dependencies, build/test/deploy workflow)
+- Design and architecture (module relationships, data flow, core entry points, conventions)
+- Implementation specifics (pipeline details, configuration patterns, file structure conventions, tool restrictions — anything too specific for the instruction file but not readable from code alone)
+
+Use the harness's file-creation primitive to create files at `.nexus/context/{chosen-name}.md`.
+
+For large projects, spawn Writer subagents per topic to generate context knowledge in parallel. Lead coordinates and reviews outputs.
+
+On completion: "context knowledge N files generated"
+
+### Step 4: Rules Initial Setup (Optional)
+
+Check whether team custom rules are needed.
+
+```
+question({ question: "Do you want to set up development rules now?", choices: [{label: "Set up", description: "Coding conventions, test policy, commit rules, etc."}, {label: Skip, description: "Can be added later via [rule] tag"}] })
+```
+
+If "Set up": present a draft based on scan results → user confirms → save via the harness's file-creation primitive to `.nexus/rules/{topic}.md`.
+
+If "Skip": inform and proceed to Step 5.
+
+### Step 5: Completion Summary
+
+Output a summary of the onboarding results.
+
+```
+## Nexus Initialization Complete
+
+### Generated Files
+- instruction file: project section — mission and essentials (<!-- PROJECT:START/END -->)
+- .nexus/context/: {list of generated files}
+- .nexus/rules/: {generated files or "none (skipped)"}
+
+### Next Steps
+- [plan] — research, analyze, and plan before execution
+- [run] — execute from a plan
+- Manual re-run trigger with `--reset` flag — re-run onboarding (existing knowledge will be backed up). See harness docs: slash_command_display.
+```