Skip to content

armgit / rea

Public Imported from GitHub
0

Reverse engineer anything with agents, from app behavior down to native binaries.

docs: show task-scoped readiness and provider preference in the quick start

Iwakurakaoru08225fd0db37d94f
diff --git a/README.md b/README.md
index 2a375d7e7..3674ab27d 100644
--- a/README.md
+++ b/README.md
@@ -95,6 +95,38 @@ MCP setup, Hopper, Ghidra, or executing the application. For a native app, confi
its engine first, then use `analyze` with that app's path. Run `doctor` when you
need diagnosis; it is not a prerequisite for each analysis.
+### Check readiness for the task at hand
+
+`rea doctor` without options is an audit of the whole integration. It checks
+every detected agent registration, the installed skill, and every optional
+analysis engine, so it can report `healthy: false` while your current task works.
+Choose a readiness scope for the work you are doing:
+
+| Task | Readiness check |
+| ----------------------------------- | -------------------------------------------------------------------------------------------- |
+| Static JavaScript/Electron analysis | None. Run `analyze-javascript-application` directly. |
+| Troubleshoot one analysis engine | `rea doctor --provider ghidra --json` (or `hopper`, `ida`) |
+| Check one agent's MCP registration | `rea doctor --client codex --json` (see [client IDs](docs/installation.md#supported-agents)) |
+| Check the installed skill | `rea doctor --skill --json` |
+
+A scoped report has `scope.mode: "explicit"`. Only `scope_checks` determine
+`healthy` and the exit status. Everything else is listed in
+`informational_checks`; you don't need to fix it for this task. For example, a
+missing engine you are not using needs no repair. `environment_healthy` still
+summarizes the full audit. `--target` adds a target check, but without a scope
+option the report remains audit-wide.
+
+When more than one installed engine supports a native target, REA does not pick
+one: opening the target fails with `code: "capability_unavailable"`,
+`details.selection_reason: "ambiguous"`, and the providers in
+`details.candidate_ids`. Choose one once: pass `--provider` on
+the CLI or `provider_id` on `open_binary`, or set `REA_ANALYSIS_PROVIDER` as a
+standing preference. An explicit selector overrides the environment variable.
+The session keeps that choice and never falls back to another engine. An
+ambiguity or unavailable-provider diagnostic asks you to choose an engine; it
+does not mean you have to install every engine. See
+[Choosing a deep-analysis provider](#choosing-a-deep-analysis-provider).
+
### Install the rea command
Install the command-line interface:
diff --git a/docs/installation.md b/docs/installation.md
index 408470534..02b7d5c94 100644
--- a/docs/installation.md
+++ b/docs/installation.md
@@ -108,7 +108,9 @@ npx -y rea-agents@latest analyze-javascript-application /absolute/path/to/app --
```
It returns the complete Evidence record directly and requires no native engine.
-Provider failures in doctor do not prevent unrelated target-free tools.
+Provider failures in doctor do not prevent unrelated target-free tools. To check
+readiness for one task instead of auditing every integration, see
+[Check readiness for the task at hand](../README.md#check-readiness-for-the-task-at-hand).
## Supported agents