🏗️ fermata: redaction-first security model, unified .botsecrets config

Realign fermata around redaction (PostToolUse) as the primary security
layer, with access control (PreToolUse) as supplementary write/bash
protection. Remove botignore.toml — policy rules now live in .botsecrets
[policy] section. Add fermata.toml as an alias for .botsecrets.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

tests/fixtures_a4.rs

@@ -9,9 +9,9 @@ fn fixture() -> TempDir {
     let root = tmp.path();
     fs::write(root.join(".botignore"), ".env\n.env.*\nconf/cert/**\nconf/mitmproxy/**\n").unwrap();
     fs::write(
-        root.join("botignore.toml"),
+        root.join(".botsecrets"),
         r#"
-[read]
+[policy.read]
 patterns = [
   "conf/localtestsettings.yaml",
   "conf/localsettings.yaml",
@@ -19,14 +19,14 @@ patterns = [
   ".claude/self-reflections/**",
 ]
 
-[write]
+[policy.write]
 patterns = [
   "conf/localtestsettings.yaml",
   "conf/localsettings.yaml",
   "conf/default-secrets.yaml",
 ]
 
-[bash]
+[policy.bash]
 deny = ["localtestsettings.yaml", "localsettings.yaml", "default-secrets.yaml", ".env"]
 ask = ["rm *", "mv *"]
 allow_prefixes = ["make test"]
@@ -103,7 +103,7 @@ fn bash_rm_somefile_asks() {
 
 #[test]
 fn read_self_reflections_asks() {
-    // Note: A.4 has self-reflections under "ask" — current toml schema uses `[read].patterns`
+    // Note: A.4 has self-reflections under "ask" — current toml schema uses `[policy.read].patterns`
     // for hard reads. This documents the gap; once toml has a `[read].ask`, switch to Ask.
     let t = fixture();
     let p = Policy::load(t.path()).unwrap();