From 591fdee72b420149c89bc5894e1bfd6fc601b33d Mon Sep 17 00:00:00 2001
From: whit3rabbit <whit3rabbit@users.noreply.github.com>
Date: Fri, 27 Mar 2026 15:30:12 -0500
Subject: [PATCH] security: configurable rate limit fail policy via
 RATE_LIMIT_FAIL_POLICY

Add "deny" as alias for "closed" policy. Change fail-closed retry-after
from 1s to 60s to give Redis time to recover. Add from_env default test.
Document env var in CLAUDE.md.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---
 CLAUDE.md                     |  1 +
 crates/proxy/src/ratelimit.rs | 22 +++++++++++++++++++---
 2 files changed, 20 insertions(+), 3 deletions(-)

diff --git a/CLAUDE.md b/CLAUDE.md
index d0cba58..7f14577 100644
--- a/CLAUDE.md
+++ b/CLAUDE.md
@@ -91,6 +91,7 @@ OPENAI_API_KEY=sk-... cargo run -p anyllm_proxy
 - `IP_ALLOWLIST`: Comma-separated CIDR ranges for IP allowlisting (e.g., `192.168.1.0/24,10.0.0.0/8`). Bare IPs also accepted. When set, only matching IPs can access the proxy.
 - `TRUST_PROXY_HEADERS`: Set to `true` or `1` to use `X-Forwarded-For` header for client IP when behind a reverse proxy. Only effective when `IP_ALLOWLIST` is set.
 - `WEBHOOK_URLS`: Comma-separated webhook URLs for request completion notifications. Fire-and-forget HTTP POST with `RequestLogEntry` JSON payload.
+- `RATE_LIMIT_FAIL_POLICY`: Behavior when Redis rate limiter is unavailable: `open` (default, allow requests) or `closed`/`deny` (reject with 503 and retry-after 60s).
 
 ### LiteLLM env var aliases
 
diff --git a/crates/proxy/src/ratelimit.rs b/crates/proxy/src/ratelimit.rs
index fb5234a..4556a2b 100644
--- a/crates/proxy/src/ratelimit.rs
+++ b/crates/proxy/src/ratelimit.rs
@@ -17,7 +17,7 @@ pub enum RateLimitFailPolicy {
 impl RateLimitFailPolicy {
     pub fn from_env_str(s: &str) -> Self {
         match s.to_lowercase().as_str() {
-            "closed" => Self::Closed,
+            "closed" | "deny" => Self::Closed,
             _ => Self::Open,
         }
     }
@@ -97,7 +97,7 @@ impl RedisRateLimiter {
                 }
                 RateLimitFailPolicy::Closed => {
                     tracing::error!(error = %e, "Redis RPM check failed, rejecting request (fail-closed)");
-                    Err(1)
+                    Err(60)
                 }
             },
         }
@@ -164,7 +164,7 @@ impl RedisRateLimiter {
                 }
                 RateLimitFailPolicy::Closed => {
                     tracing::error!(error = %e, "Redis TPM check failed, rejecting request (fail-closed)");
-                    Err(1)
+                    Err(60)
                 }
             },
         }
@@ -265,9 +265,25 @@ mod tests {
             RateLimitFailPolicy::from_env_str("CLOSED"),
             RateLimitFailPolicy::Closed
         ));
+        assert!(matches!(
+            RateLimitFailPolicy::from_env_str("deny"),
+            RateLimitFailPolicy::Closed
+        ));
+        assert!(matches!(
+            RateLimitFailPolicy::from_env_str("DENY"),
+            RateLimitFailPolicy::Closed
+        ));
         assert!(matches!(
             RateLimitFailPolicy::from_env_str("unknown"),
             RateLimitFailPolicy::Open
         ));
     }
+
+    #[test]
+    fn fail_policy_defaults_to_open() {
+        // When RATE_LIMIT_FAIL_POLICY is unset, from_env should return Open.
+        std::env::remove_var("RATE_LIMIT_FAIL_POLICY");
+        let policy = RateLimitFailPolicy::from_env();
+        assert_eq!(policy, RateLimitFailPolicy::Open);
+    }
 }