@a5c-ai/babysitter-github 0.1.1-staging.0a3fc67d

package/skills/babysit/SKILL.md

@@ -0,0 +1,77 @@
+---
+name: babysit
+description: >-
+  Orchestrate via @babysitter. Use this skill when asked to babysit a run,
+  orchestrate a process or whenever it is called explicitly. (babysit,
+  babysitter, orchestrate, orchestrate a run, workflow, etc.)
+---
+
+# babysit
+
+Orchestrate `.a5c/runs/<runId>/` through iterative execution.
+
+## Dependencies
+
+### Babysitter SDK and CLI
+
+Read the SDK version from `versions.json` to ensure version compatibility:
+
+```bash
+SDK_VERSION=$(node -e "try{console.log(JSON.parse(require('fs').readFileSync('${PLUGIN_ROOT}/versions.json','utf8')).sdkVersion||'latest')}catch{console.log('latest')}")
+sudo npm i -g @a5c-ai/babysitter-sdk@$SDK_VERSION
+# sudo is depending on the env and system
+```
+
+then use the CLI alias: CLI="babysitter"
+
+**Alternatively, use the CLI alias:** `CLI="npx -y @a5c-ai/babysitter-sdk@$SDK_VERSION"`
+
+### jq
+
+make sure you have jq installed and available in the path. if not, install it.
+
+## Instructions
+
+Run the following command to get full orchestration instructions:
+
+```bash
+babysitter instructions:babysit-skill --harness github-copilot --json
+```
+
+Follow the instructions returned by the command above to orchestrate the run.
+
+## GitHub Copilot CLI -- In-Turn Loop Model
+
+**IMPORTANT**: Unlike Claude Code, GitHub Copilot CLI does NOT have a Stop hook
+that can drive the orchestration loop between turns. The `sessionEnd` hook output
+is ignored and cannot block or restart the session.
+
+Therefore, you MUST use **in-turn iteration**: run the full orchestration loop
+within a single session turn. The pattern is:
+
+1. `babysitter run:iterate --json` -- get pending actions
+2. For each pending action: execute it (run tasks, post results via `task:post`)
+3. `babysitter run:iterate --json` -- check for more pending actions
+4. Repeat steps 2-3 until run completes or reaches a breakpoint requiring user input
+5. If a breakpoint requires user input, ask the user and post the response, then continue iterating
+
+All iteration happens within the same turn -- do NOT rely on hooks to re-enter
+the orchestration loop. The agent drives the loop directly by calling
+`run:iterate` repeatedly until completion.
+
+### Loop Example
+
+```bash
+# Initial iterate
+RESULT=$(babysitter run:iterate --run-id "$RUN_ID" --json)
+STATUS=$(echo "$RESULT" | jq -r '.status')
+
+while [ "$STATUS" != "completed" ] && [ "$STATUS" != "failed" ]; do
+  # Process pending actions from RESULT
+  # ... execute tasks, post results ...
+
+  # Iterate again
+  RESULT=$(babysitter run:iterate --run-id "$RUN_ID" --json)
+  STATUS=$(echo "$RESULT" | jq -r '.status')
+done
+```

package/skills/call/SKILL.md

@@ -0,0 +1,17 @@
+---
+name: call
+description: Start a new Babysitter orchestration run from GitHub Copilot CLI.
+---
+
+# call
+
+Load and use the installed `babysit` skill.
+
+Resolve the request in `call` mode:
+
+- treat everything after `$call` as the initial Babysitter request for a new
+  orchestration run
+- create the process, create the run, and enter the Babysitter loop
+- using this always means the user meant an interactive run
+- do not create a separate command surface here; this skill only forwards into
+  `babysit`

package/skills/doctor/SKILL.md

@@ -0,0 +1,16 @@
+---
+name: doctor
+description: Diagnose Babysitter run health, integration issues, or orchestration failures.
+---
+
+# doctor
+
+Load and use the installed `babysit` skill.
+
+Resolve the request in `doctor` mode:
+
+- treat everything after `$doctor` as the run selector, subsystem, or problem
+  statement to diagnose
+- focus on run health, orchestration correctness, and recovery guidance
+- do not create a separate command surface here; this skill only forwards into
+  `babysit`

package/skills/help/SKILL.md

@@ -0,0 +1,15 @@
+---
+name: help
+description: Explain Babysitter GitHub Copilot CLI usage, modes, setup, or operational behavior.
+---
+
+# help
+
+Load and use the installed `babysit` skill.
+
+Resolve the request in `help` mode:
+
+- treat everything after `$help` as the help target or question
+- focus on explaining the right Babysitter flow or integration surface
+- do not create a separate command surface here; this skill only forwards into
+  `babysit`

package/skills/observe/SKILL.md

@@ -0,0 +1,15 @@
+---
+name: observe
+description: Observe, inspect, or monitor a Babysitter run.
+---
+
+# observe
+
+Load and use the installed `babysit` skill.
+
+Resolve the request in `observe` mode:
+
+- treat everything after `$observe` as the run selector or observation target
+- focus on inspection, monitoring, and state visibility
+- do not create a separate command surface here; this skill only forwards into
+  `babysit`

package/skills/plan/SKILL.md

@@ -0,0 +1,16 @@
+---
+name: plan
+description: Plan a Babysitter workflow without executing the run.
+---
+
+# plan
+
+Load and use the installed `babysit` skill.
+
+Resolve the request in `plan` mode:
+
+- treat everything after `$plan` as the planning request
+- focus on building the best process possible without creating and running the
+  actual run unless the user explicitly changes mode
+- do not create a separate command surface here; this skill only forwards into
+  `babysit`

package/skills/resume/SKILL.md

@@ -0,0 +1,15 @@
+---
+name: resume
+description: Resume an existing Babysitter run from GitHub Copilot CLI.
+---
+
+# resume
+
+Load and use the installed `babysit` skill.
+
+Resolve the request in `resume` mode:
+
+- treat everything after `$resume` as the run selector or run id
+- focus on restoring the orchestration context and continuing the run honestly
+- do not create a separate command surface here; this skill only forwards into
+  `babysit`

package/skills/retrospect/SKILL.md

@@ -0,0 +1,55 @@
+---
+name: retrospect
+description: Summarize or retrospect on one or more completed Babysitter runs.
+argument-hint: "[run-id...] [--all] Optional run IDs or --all for all runs"
+---
+
+# retrospect
+
+Load and use the installed `babysit` skill.
+
+Resolve the request in `retrospect` mode:
+
+- treat everything after `$retrospect` as the run selector(s) to summarize
+- focus on the run history, outcomes, lessons, and gaps
+- do not create a separate command surface here; this skill only forwards into
+  `babysit`
+
+## Phase 1: Resolve Target Run(s)
+
+- If `--all` or "all" is present in args: list all runs via `ls -lt .a5c/runs/` and collect all completed/failed run IDs
+- If multiple run IDs are provided: use all of them
+- Otherwise: existing behavior (resolve the latest single run)
+- Use interactive prompts to confirm run selection when possible
+
+## Phase 2: Load Run Data
+
+For each selected run, load:
+- `run.json` metadata
+- Journal events
+- Task definitions and results
+- State snapshots
+
+## Phase 3: Analysis
+
+Perform standard per-run analysis (outcomes, process effectiveness, suggestions).
+
+### Cross-Run Pattern Analysis (multi-run mode)
+
+When analyzing multiple runs, additionally cover:
+- **Common failure modes** across runs
+- **Velocity trends** (tasks/time across runs)
+- **Process evolution** (how processes changed over time)
+- **Repeated breakpoint patterns**
+
+## Phase 4: Suggestions
+
+Provide actionable suggestions for process improvements, optimizations, and fixes.
+
+## Phase 5: Implementation
+
+If the user agrees, implement improvements to processes, skills, or configuration.
+
+## Phase 6: Cleanup Suggestion
+
+After analysis, suggest: "Consider running `babysitter cleanup` (or `/babysitter:cleanup`) to clean up old run data and reclaim disk space."

package/skills/user-install/SKILL.md

@@ -0,0 +1,15 @@
+---
+name: user-install
+description: Install the user-level Babysitter GitHub Copilot CLI setup.
+---
+
+# user-install
+
+Load and use the installed `babysit` skill.
+
+Resolve the request in `user-install` mode:
+
+- treat everything after `$user-install` as the user-setup request
+- focus on user profile, user-level install, and personal defaults
+- do not create a separate command surface here; this skill only forwards into
+  `babysit`

package/versions.json

@@ -0,0 +1,3 @@
+{
+  "sdkVersion": "0.0.183-staging.0a3fc67d"
+}