loki-mode 6.11.0 → 6.11.1

package/SKILL.md

@@ -3,7 +3,7 @@ name: loki-mode
 description: Multi-agent autonomous startup system. Triggers on "Loki Mode". Takes PRD to deployed product with minimal human intervention. Requires --dangerously-skip-permissions flag.
 ---
 
-# Loki Mode v6.11.0
+# Loki Mode v6.11.1
 
 **You are an autonomous agent. You make decisions. You do not ask questions. You do not stop.**
 
@@ -267,4 +267,4 @@ The following features are documented in skill modules but not yet fully automat
 | Quality gates 3-reviewer system | Implemented (v5.35.0) | 5 specialist reviewers in `skills/quality-gates.md`; execution in run.sh |
 | Benchmarks (HumanEval, SWE-bench) | Infrastructure only | Runner scripts and datasets exist in `benchmarks/`; no published results |
 
-**v6.11.0 | [Autonomi](https://www.autonomi.dev/) flagship product | ~260 lines core**
+**v6.11.1 | [Autonomi](https://www.autonomi.dev/) flagship product | ~260 lines core**

package/VERSION

@@ -1 +1 @@
-6.11.0
+6.11.1

package/autonomy/loki

@@ -11585,7 +11585,22 @@ cmd_remote() {
     emit_event session cli remote_start "prd=${prd_abs:-none}"
 
     # Hand off to claude remote-control
-    exec claude remote-control ${rc_flags[@]+"${rc_flags[@]}"}
+    # Note: not using exec so we can provide guidance if it fails
+    claude remote-control ${rc_flags[@]+"${rc_flags[@]}"}
+    local rc_exit=$?
+
+    if [ $rc_exit -ne 0 ]; then
+        echo ""
+        echo -e "${YELLOW}Troubleshooting:${NC}"
+        echo ""
+        echo -e "  ${CYAN}Workspace not trusted?${NC}"
+        echo "    Run 'claude' in this directory first to accept the trust dialog,"
+        echo "    then exit (/exit) and run 'loki remote' again."
+        echo ""
+        echo -e "  ${CYAN}Not logged in?${NC}"
+        echo "    Run 'claude' then '/login' -- requires a Pro or Max subscription."
+        exit $rc_exit
+    fi
 }
 
 # Syslog/SIEM integration management

package/dashboard/__init__.py

@@ -7,7 +7,7 @@ Modules:
     control: Session control API (start/stop/pause/resume)
 """
 
-__version__ = "6.11.0"
+__version__ = "6.11.1"
 
 # Expose the control app for easy import
 try:

package/docs/INSTALLATION.md

@@ -2,7 +2,7 @@
 
 The flagship product of [Autonomi](https://www.autonomi.dev/). Complete installation instructions for all platforms and use cases.
 
-**Version:** v6.11.0
+**Version:** v6.11.1
 
 ---
 

package/mcp/__init__.py

@@ -57,4 +57,4 @@ try:
 except ImportError:
     __all__ = ['mcp']
 
-__version__ = '6.11.0'
+__version__ = '6.11.1'

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "loki-mode",
-  "version": "6.11.0",
+  "version": "6.11.1",
   "description": "Loki Mode by Autonomi - Multi-agent autonomous startup system for Claude Code, Codex CLI, and Gemini CLI",
   "keywords": [
     "agent",

package/skills/openspec-integration.md

@@ -9,7 +9,7 @@
 - Your prompt contains an `OPENSPEC DELTA CONTEXT` section
 - The project has `.loki/openspec/delta-context.json`
 - The session was started with `--openspec` flag
-- Tasks in `.loki/queue/pending.json` have `openspec_group` metadata
+- Tasks in `.loki/queue/pending.json` have `metadata.openspec_group`
 
 **If none of the above are true, do not load this module.**
 
@@ -24,7 +24,7 @@ New behavior that does not exist in the codebase yet.
 1. Create NEW files and functions following existing codebase patterns
 2. Do NOT modify existing code unless the new feature integrates with it
 3. Write tests for every scenario (GIVEN/WHEN/THEN from the delta spec)
-4. Reference: `delta-context.json` entries with `"type": "added"`
+4. Reference: `delta-context.json` → `deltas.<domain>.added[]`
 
 ```
 # Mental model for ADDED
@@ -39,7 +39,7 @@ Existing behavior that is changing. This is the most common delta type in brownf
 2. Modify IN PLACE -- do NOT create new files for modified behavior
 3. Check the `(Previously: ...)` annotation to understand what changed
 4. Update existing tests to match the new behavior
-5. Reference: `delta-context.json` entries with `"type": "modified"`
+5. Reference: `delta-context.json` → `deltas.<domain>.modified[]` (each entry has a `"previously"` field)
 
 ```
 # Mental model for MODIFIED
@@ -56,7 +56,7 @@ Behavior that is being deprecated or deleted.
 2. Check the `(Deprecated: ...)` annotation for the reason
 3. Remove associated tests
 4. Ensure no orphaned imports or dead code remains
-5. Reference: `delta-context.json` entries with `"type": "removed"`
+5. Reference: `delta-context.json` → `deltas.<domain>.removed[]` (each entry has a `"reason"` field)
 
 ```
 # Mental model for REMOVED
@@ -69,18 +69,23 @@ Read "(Deprecated: ...)" -> Find existing code -> Remove code -> Remove tests ->
 
 Tasks are generated from OpenSpec `tasks.md` and loaded into `.loki/queue/pending.json`.
 
-- Each task has `openspec_group` metadata indicating its task group number
+- Each task has `metadata.openspec_group` indicating its group name (string, from `## N. Group Name` headings in tasks.md)
 - Execute tasks in group order (group 1 before group 2, etc.)
 - Within a group, tasks can run in parallel if they touch different files
 - Mark tasks complete in the queue when done
 
 ```json
 {
-  "id": "task-3",
+  "id": "openspec-1.3",
   "title": "Implement session timeout change",
-  "openspec_group": 1,
-  "delta_type": "modified",
-  "spec_ref": "auth/spec.md#session-expiration"
+  "description": "[OpenSpec] Authentication: Implement session timeout change",
+  "priority": "medium",
+  "status": "pending",
+  "source": "openspec",
+  "metadata": {
+    "openspec_source": "tasks.md",
+    "openspec_group": "Authentication"
+  }
 }
 ```
 
@@ -93,7 +98,7 @@ After implementing a requirement, verify its scenarios.
 1. Each scenario has GIVEN (precondition), WHEN (action), THEN (expected outcome)
 2. Write test cases that map 1:1 to scenarios
 3. Use the scenario name as the test name for traceability
-4. Verification results are tracked in `.loki/openspec/verification-results.json`
+4. Verification results are tracked in `.loki/openspec/verification-map.json`
 
 ```python
 # Scenario: "Idle timeout" -> test name matches scenario
@@ -106,19 +111,43 @@ def test_idle_timeout():
     assert session.is_expired()
 ```
 
+The verification map tracks each scenario with `"verified": false` initially. After tests pass, the orchestrator updates verified status.
+
+```json
+{
+  "scenarios": [
+    {
+      "domain": "auth",
+      "requirement": "Session Expiration",
+      "scenario": "Idle timeout",
+      "given": "an authenticated session",
+      "when": "15 minutes pass without activity",
+      "then": "the session is invalidated",
+      "verified": false
+    }
+  ]
+}
+```
+
 ---
 
 ## Source Mapping
 
-`.loki/openspec/source-map.json` maps each task ID to its origin in the spec files.
+`.loki/openspec/source-map.json` maps each task ID to its origin in tasks.md.
 
 | Field | Purpose |
 |-------|---------|
-| `task_id` | Queue task identifier |
-| `spec_file` | Source spec file path |
-| `requirement` | Requirement name |
-| `scenario` | Scenario name (if applicable) |
-| `line` | Line number in spec file |
+| `file` | Source file (always `tasks.md`) |
+| `line` | Line number in tasks.md (1-indexed) |
+| `group` | Task group name from `## N. Group Name` heading |
+
+```json
+{
+  "openspec-1.1": { "file": "tasks.md", "line": 3, "group": "Authentication" },
+  "openspec-1.2": { "file": "tasks.md", "line": 4, "group": "Authentication" },
+  "openspec-2.1": { "file": "tasks.md", "line": 7, "group": "Dashboard" }
+}
+```
 
 Use this to trace implementation decisions back to the specification.
 
@@ -126,12 +155,28 @@ Use this to trace implementation decisions back to the specification.
 
 ## Complexity Levels
 
-| Level | Tasks | Spec Files | Design | Agent Strategy |
-|-------|-------|------------|--------|----------------|
-| simple | 1-3 | 1 | none | Single agent, sequential |
-| standard | 4-10 | 2-5 | present | Parallel where possible |
-| complex | 11-20 | 5-10 | present | Task tool parallelization |
-| enterprise | 20+ | 10+ | present | Full agent team |
+The adapter classifies complexity based on task count, spec file count, and design.md presence. Checked in order (first match wins):
+
+| Level | Condition | Agent Strategy |
+|-------|-----------|----------------|
+| enterprise | 20+ tasks OR 10+ spec files | Full agent team |
+| complex | 11-20 tasks OR 6-10 spec files | Task tool parallelization |
+| standard | 4-10 tasks OR 2-5 spec files OR design.md present | Parallel where possible |
+| simple | 1-3 tasks, 1 spec file, no design | Single agent, sequential |
+
+---
+
+## Output Files Reference
+
+All output files are written to `.loki/` by the adapter (`autonomy/openspec-adapter.py`):
+
+| File | Purpose |
+|------|---------|
+| `.loki/openspec-prd-normalized.md` | Synthesized PRD from proposal + specs + design |
+| `.loki/openspec-tasks.json` | Task list parsed from tasks.md |
+| `.loki/openspec/delta-context.json` | Structured deltas with stats and complexity |
+| `.loki/openspec/source-map.json` | Task ID to tasks.md line mapping |
+| `.loki/openspec/verification-map.json` | Scenario verification tracking |
 
 ---