AGENTS.md

1# Agent Instructions
2 
3This project uses **bd** (beads) for issue tracking. Run `bd prime` for full workflow context.
4 
5## Quick Reference
6 
7```bash
8bd ready              # Find available work
9bd show <id>          # View issue details
10bd update <id> --claim  # Claim work atomically
11bd close <id>         # Complete work
12bd dolt push          # Push beads data to remote
13```
14 
15## Non-Interactive Shell Commands
16 
17**ALWAYS use non-interactive flags** with file operations to avoid hanging on confirmation prompts.
18 
19Shell commands like `cp`, `mv`, and `rm` may be aliased to include `-i` (interactive) mode on some systems, causing the agent to hang indefinitely waiting for y/n input.
20 
21**Use these forms instead:**
22```bash
23# Force overwrite without prompting
24cp -f source dest           # NOT: cp source dest
25mv -f source dest           # NOT: mv source dest
26rm -f file                  # NOT: rm file
27 
28# For recursive operations
29rm -rf directory            # NOT: rm -r directory
30cp -rf source dest          # NOT: cp -r source dest
31```
32 
33**Other commands that may prompt:**
34- `scp` - use `-o BatchMode=yes` for non-interactive
35- `ssh` - use `-o BatchMode=yes` to fail instead of prompting
36- `apt-get` - use `-y` flag
37- `brew` - use `HOMEBREW_NO_AUTO_UPDATE=1` env var
38 
39<!-- BEGIN BEADS INTEGRATION v:1 profile:minimal hash:ca08a54f -->
40## Beads Issue Tracker
41 
42This project uses **bd (beads)** for issue tracking. Run `bd prime` to see full workflow context and commands.
43 
44### Quick Reference
45 
46```bash
47bd ready              # Find available work
48bd show &lt;id&gt;          # View issue details
49bd update &lt;id&gt; --claim  # Claim work
50bd close &lt;id&gt;         # Complete work
51```
52 
53### Rules
54 
55- Use `bd` for ALL task tracking — do NOT use TodoWrite, TaskCreate, or markdown TODO lists
56- Run `bd prime` for detailed command reference and session close protocol
57- Use `bd remember` for persistent knowledge — do NOT use MEMORY.md files
58 
59## Session Completion
60 
61**When ending a work session**, you MUST complete ALL steps below. Work is NOT complete until `git push` succeeds.
62 
63**MANDATORY WORKFLOW:**
64 
651. **File issues for remaining work** - Create issues for anything that needs follow-up
662. **Run quality gates** (if code changed) - Tests, linters, builds
673. **Update issue status** - Close finished work, update in-progress items
684. **PUSH TO REMOTE** - This is MANDATORY:
69```bash
70   git pull --rebase
71   bd dolt push
72   git push
73   git status  # MUST show "up to date with origin"
74```
755. **Clean up** - Clear stashes, prune remote branches
766. **Verify** - All changes committed AND pushed
777. **Hand off** - Provide context for next session
78 
79**CRITICAL RULES:**
80- Work is NOT complete until `git push` succeeds
81- NEVER stop before pushing - that leaves work stranded locally
82- NEVER say "ready to push when you are" - YOU must push
83- If push fails, resolve and retry until it succeeds
84<!-- END BEADS INTEGRATION -->
85 
86<!-- code-review-graph MCP tools -->
87## MCP Tools: code-review-graph
88 
89**IMPORTANT: This project has a knowledge graph. ALWAYS use the
90code-review-graph MCP tools BEFORE using Grep/Glob/Read to explore
91the codebase.** The graph is faster, cheaper (fewer tokens), and gives
92you structural context (callers, dependents, test coverage) that file
93scanning cannot.
94 
95### When to use graph tools FIRST
96 
97- **Exploring code**: `semantic_search_nodes_tool` or `query_graph_tool` instead of Grep
98- **Understanding impact**: `get_impact_radius_tool` instead of manually tracing imports
99- **Code review**: `detect_changes_tool` + `get_review_context_tool` instead of reading entire files
100- **Finding relationships**: `query_graph_tool` with callers_of/callees_of/imports_of/tests_for
101- **Architecture questions**: `get_architecture_overview_tool` + `list_communities_tool`
102 
103Fall back to Grep/Glob/Read **only** when the graph doesn't cover what you need.
104 
105### Key Tools
106 
107| Tool | Use when |
108|------|----------|
109| `detect_changes_tool` | Reviewing code changes — gives risk-scored analysis |
110| `get_review_context_tool` | Need source snippets for review — token-efficient |
111| `get_impact_radius_tool` | Understanding blast radius of a change |
112| `get_affected_flows_tool` | Finding which execution paths are impacted |
113| `query_graph_tool` | Tracing callers, callees, imports, tests, dependencies |
114| `semantic_search_nodes_tool` | Finding functions/classes by name or keyword |
115| `get_architecture_overview_tool` | Understanding high-level codebase structure |
116| `refactor_tool` | Planning renames, finding dead code |
117 
118### Workflow
119 
1201. The graph auto-updates on file changes (via hooks).
1212. Use `detect_changes_tool` for code review.
1223. Use `get_affected_flows_tool` to understand impact.
1234. Use `query_graph_tool` pattern="tests_for" to check coverage.
124