Solution for long term memory for agent coding CLIs and to facilitate handoff between different agent vendors
Rust
docs(config): describe both evaluation gate modes
AkitaOnRails06e6fdd179f7
diff --git a/crates/ai-memory-cli/templates/config.default.toml b/crates/ai-memory-cli/templates/config.default.tomlindex 464df6c6..3ee4726a 100644--- a/crates/ai-memory-cli/templates/config.default.toml+++ b/crates/ai-memory-cli/templates/config.default.toml@@ -518,13 +518,14 @@ pending_path = "_pending/auto-improve"# max_rejection_context = 50# rejection_context_days = 180-# Optional executable evaluation gate for selected proposal targets. Disabled by-# default; operators supply the command and scoring policy. The command is run-# directly (not through a shell), split on whitespace, with JSON on stdin. Use a-# wrapper script for paths with spaces, custom environment, or multiple commands.-# It must print JSON such as { "score_before": 0.72, "score_after": 0.76, "passed": true }.-# Or set builtin = "structure" to run the built-in structural checks in process-# instead; builtin and command are mutually exclusive.+# Optional evaluation gate for selected proposal targets. Disabled by default;+# choose either the built-in structural scorer or an external command. A command+# is run directly (not through a shell), split on whitespace, with JSON on stdin.+# Use a wrapper script for paths with spaces, custom environment, or multiple+# commands. It must print JSON such as+# { "score_before": 0.72, "score_after": 0.76, "passed": true }.+# Set builtin = "structure" to run the built-in checks in process instead;+# builtin and command are mutually exclusive.[auto_improve.eval]enabled = falsecommand = ""diff --git a/docs/config-properties.md b/docs/config-properties.mdindex c669c1f1..166a1308 100644--- a/docs/config-properties.md+++ b/docs/config-properties.md@@ -216,7 +216,7 @@ The defaults named `DEFAULT_AUTO_IMPROVE_*` are constants in `ai-memory-consolid| Key | Default | Env equivalent | Purpose || --- | --- | --- | --- |-| `enabled` | `false` | `AI_MEMORY_AUTO_IMPROVE__EVAL__ENABLED` | Turn on the external eval command gate. |+| `enabled` | `false` | `AI_MEMORY_AUTO_IMPROVE__EVAL__ENABLED` | Turn on the proposal evaluation gate. || `command` | empty | `AI_MEMORY_AUTO_IMPROVE__EVAL__COMMAND` | Executable plus whitespace-separated arguments, run directly without a shell. || `builtin` | empty | `AI_MEMORY_AUTO_IMPROVE__EVAL__BUILTIN` | Scorer run in process instead of `command`. The only value is `structure`. Set `builtin` or `command`, not both. || `timeout_secs` | `120` | `AI_MEMORY_AUTO_IMPROVE__EVAL__TIMEOUT_SECS` | Timeout per proposal eval command. |