RepoPrompt Tools Reminder
Continue your current workflow using RepoPrompt MCP tools instead of built-in alternatives.
File & Code
| Task |
Use |
Not |
| Search paths/content |
file_search |
grep, find, Glob |
| Read file (whole or sliced) |
read_file |
cat, head, Read |
| Directory tree |
get_file_tree |
ls, find |
| Signatures / overview |
get_code_structure |
reading whole files |
| Edit file |
apply_edits |
sed, Edit |
| Create / delete / move |
file_actions |
touch, rm, mv, Write |
| Git status / diff / log / blame |
git |
shelling out for analysis |
Context & Planning
| Tool |
Use for |
manage_selection |
Curate the file set used by chat, builder, and exports. Refresh before each planning call. Modes: full, slices, codemap_only. |
workspace_context |
Snapshot current prompt + selection + token budget; also exports. |
prompt |
Read/set the shared prompt; list or select copy presets. |
context_builder |
Heavy discovery sub-agent — describe the task, it curates files + rewrites the prompt. response_type: clarify / plan / question / review. Pass export_response:true to hand the result to a child agent. |
ask_oracle |
Chat-mode reasoning over the current selection. Continue existing chats (new_chat:false) rather than opening new ones. Modes: chat / plan / review. |
oracle_chat_log |
Recover recent Oracle messages after compaction. |
ask_user |
Ask the user when ambiguity is load-bearing — don't guess at requirements. |
Agent Delegation -- multi_agent_v2
Dispatch a sub-agent when a side investigation or delegated chunk of work would otherwise flood this session's context. Use shipped Codex agents directly, not RepoPrompt role labels. Always pass fork_turns:"none" because these agents read context from the first message. Always pass the listed reasoning_effort.
| Agent type |
Reasoning |
Use |
code_mapper |
low |
Fast read-only probes, git archaeology, wiring lookups, narrow in-repo checks. One question per probe. |
docs_researcher |
low |
External docs, library/API behavior, web research, standards. |
implementation_engineer |
low |
Well-scoped implementation work from an existing plan. |
implementation_lead |
high |
Complex implementation work, architecture-sensitive changes, or branch-level ownership. |
investigation_lead |
high |
Multi-step root cause analysis and evidence gathering. |
test_automator |
medium |
Focused test, benchmark, or verification work. |
browser_debugger |
medium |
Browser/UI runtime debugging. |
workflow_orchestrator |
xhigh |
Plan critique, dependency/risk analysis, coordination review. |
Key operations:
- Start:
spawn_agent with task_name, agent_type, reasoning_effort, fork_turns:"none", and a self-contained message.
- Wait/check:
wait_agent, then list_agents for current status.
- Continue:
followup_task with the target task name.
- Close:
close_agent once you have recorded the result.
Fan-out pattern: call spawn_agent once per independent probe, then use wait_agent and list_agents until every spawned agent has returned. There is no detach flag in this workflow; spawn_agent already creates independent agents.
Export handoff: when context_builder or ask_oracle returns oracle_export_path, include that path inside the child agent's next message so it reads the export with read_file.
Quick Reference
// Search · Read · Edit · File ops
{"tool":"file_search","args":{"pattern":"keyword","mode":"auto"}}
{"tool":"read_file","args":{"path":"Root/file.swift","start_line":50,"limit":30}}
{"tool":"apply_edits","args":{"path":"Root/file.swift","search":"old","replace":"new"}}
{"tool":"file_actions","args":{"action":"create","path":"Root/new.swift","content":"..."}}
// Selection · Builder · Oracle
{"tool":"manage_selection","args":{"op":"add","paths":["Root/path/file.swift"]}}
{"tool":"context_builder","args":{"instructions":"<task>","response_type":"plan"}}
{"tool":"ask_oracle","args":{"message":"...","mode":"plan","new_chat":false}}
// Delegate · Fan-out · Steer · Cleanup
{"tool":"spawn_agent","args":{"task_name":"probe_x","agent_type":"code_mapper","reasoning_effort":"low","fork_turns":"none","message":"<question>"}}
{"tool":"wait_agent","args":{"timeout_ms":60000}}
{"tool":"list_agents","args":{}}
{"tool":"followup_task","args":{"target":"probe_x","message":"now do Y"}}
{"tool":"close_agent","args":{"target":"probe_x"}}
Continue with your task using these tools.
1---2name: rp-reminder-v23description: Reminder to use RepoPrompt MCP tools4---56# RepoPrompt Tools Reminder78Continue your current workflow using RepoPrompt MCP tools instead of built-in alternatives.910## File & Code1112| Task | Use | Not |13|------|-----|-----|14| Search paths/content | `file_search` | grep, find, Glob |15| Read file (whole or sliced) | `read_file` | cat, head, Read |16| Directory tree | `get_file_tree` | ls, find |17| Signatures / overview | `get_code_structure` | reading whole files |18| Edit file | `apply_edits` | sed, Edit |19| Create / delete / move | `file_actions` | touch, rm, mv, Write |20| Git status / diff / log / blame | `git` | shelling out for analysis |2122## Context & Planning2324| Tool | Use for |25|------|---------|26| `manage_selection` | Curate the file set used by chat, builder, and exports. Refresh before each planning call. Modes: `full`, `slices`, `codemap_only`. |27| `workspace_context` | Snapshot current prompt + selection + token budget; also exports. |28| `prompt` | Read/set the shared prompt; list or select copy presets. |29| `context_builder` | Heavy discovery sub-agent — describe the task, it curates files + rewrites the prompt. `response_type`: `clarify` / `plan` / `question` / `review`. Pass `export_response:true` to hand the result to a child agent. |30| `ask_oracle` | Chat-mode reasoning over the current selection. Continue existing chats (`new_chat:false`) rather than opening new ones. Modes: `chat` / `plan` / `review`. |31| `oracle_chat_log` | Recover recent Oracle messages after compaction. |32| `ask_user` | Ask the user when ambiguity is load-bearing — don't guess at requirements. |3334## Agent Delegation -- `multi_agent_v2`3536Dispatch a sub-agent when a side investigation or delegated chunk of work would otherwise flood this session's context. Use shipped Codex agents directly, not RepoPrompt role labels. Always pass `fork_turns:"none"` because these agents read context from the first message. Always pass the listed `reasoning_effort`.3738| Agent type | Reasoning | Use |39|------------|-----------|-----|40| `code_mapper` | `low` | Fast read-only probes, git archaeology, wiring lookups, narrow in-repo checks. One question per probe. |41| `docs_researcher` | `low` | External docs, library/API behavior, web research, standards. |42| `implementation_engineer` | `low` | Well-scoped implementation work from an existing plan. |43| `implementation_lead` | `high` | Complex implementation work, architecture-sensitive changes, or branch-level ownership. |44| `investigation_lead` | `high` | Multi-step root cause analysis and evidence gathering. |45| `test_automator` | `medium` | Focused test, benchmark, or verification work. |46| `browser_debugger` | `medium` | Browser/UI runtime debugging. |47| `workflow_orchestrator` | `xhigh` | Plan critique, dependency/risk analysis, coordination review. |4849**Key operations:**50- Start: `spawn_agent` with `task_name`, `agent_type`, `reasoning_effort`, `fork_turns:"none"`, and a self-contained `message`.51- Wait/check: `wait_agent`, then `list_agents` for current status.52- Continue: `followup_task` with the target task name.53- Close: `close_agent` once you have recorded the result.5455**Fan-out pattern:** call `spawn_agent` once per independent probe, then use `wait_agent` and `list_agents` until every spawned agent has returned. There is no detach flag in this workflow; `spawn_agent` already creates independent agents.5657**Export handoff:** when `context_builder` or `ask_oracle` returns `oracle_export_path`, include that path inside the child agent's next `message` so it reads the export with `read_file`.5859## Quick Reference6061```json62// Search · Read · Edit · File ops63{"tool":"file_search","args":{"pattern":"keyword","mode":"auto"}}64{"tool":"read_file","args":{"path":"Root/file.swift","start_line":50,"limit":30}}65{"tool":"apply_edits","args":{"path":"Root/file.swift","search":"old","replace":"new"}}66{"tool":"file_actions","args":{"action":"create","path":"Root/new.swift","content":"..."}}6768// Selection · Builder · Oracle69{"tool":"manage_selection","args":{"op":"add","paths":["Root/path/file.swift"]}}70{"tool":"context_builder","args":{"instructions":"<task>","response_type":"plan"}}71{"tool":"ask_oracle","args":{"message":"...","mode":"plan","new_chat":false}}7273// Delegate · Fan-out · Steer · Cleanup74{"tool":"spawn_agent","args":{"task_name":"probe_x","agent_type":"code_mapper","reasoning_effort":"low","fork_turns":"none","message":"<question>"}}75{"tool":"wait_agent","args":{"timeout_ms":60000}}76{"tool":"list_agents","args":{}}77{"tool":"followup_task","args":{"target":"probe_x","message":"now do Y"}}78{"tool":"close_agent","args":{"target":"probe_x"}}79```8081Continue with your task using these tools.
novotnyllc