```text
██████╗ ██╗ ██╗██████╗ ███╗ ██╗██████╗ █████╗ ████████╗███████╗
██╔══██╗██║ ██║██╔══██╗████╗ ██║██╔══██╗██╔══██╗╚══██╔══╝██╔════╝
██████╔╝██║ ██║██████╔╝██╔██╗ ██║██████╔╝███████║ ██║ █████╗
██╔══██╗██║ ██║██╔══██╗██║╚██╗██║██╔══██╗██╔══██║ ██║ ██╔══╝
██████╔╝╚██████╔╝██║ ██║██║ ╚████║██║ ██║██║ ██║ ██║ ███████╗
╚═════╝ ╚═════╝ ╚═╝ ╚═╝╚═╝ ╚═══╝╚═╝ ╚═╝╚═╝ ╚═╝ ╚═╝ ╚══════╝
```
### **Terminal HUD, Hard Spend Caps & Cache Guard for Coding Agents**
[](https://github.com/ronitgupta138/BurnRate/stargazers)
[](https://crates.io/crates/burnrate-agent)
[](LICENSE)
[](https://www.rust-lang.org)
[](#-verification--benchmarks)
[](https://hits.sh/github.com/ronitgupta138/burnrate/)
[](https://github.com/sponsors/ronitgupta138)
**Never wake up to an unexpected $40 Claude Code or Cursor bill again.**
A lightweight, zero-overhead reverse proxy and CLI HUD that gives developers absolute cost control and prompt-cache leak alerts over autonomous AI sessions.
---
## 🎯 The Problem: Why Developers Fear Agent Bills
Software engineers adopting autonomous terminal coding assistants (Claude Code, Cursor, Codex, OpenCode, Aider) frequently suffer from **bill anxiety**:
1. **Silent Infinite Retry Loops:** An agent gets confused by a compiler or test error and loops 35 times, burning $15 in 4 minutes.
2. **Context Explosions:** The agent inadvertently reads `package-lock.json`, `dist/`, or a 10MB SQLite database into context.
3. **Prompt Cache Breakage:** A tiny dynamic timestamp or un-sorted JSON injected at the top of a prompt secretly invalidates the 90% prompt cache discount, multiplying token costs by 10x with zero warning.
`burnrate` acts as an in-line cost shield: a zero-latency reverse proxy and terminal HUD that monitors session spend and terminates runaway loops before they drain your wallet.
---
## 🖥️ Live Terminal Status HUD
Renders directly to stderr with live cost consumption, token burn rates, and prompt-cache health:
```text
┌─────────────────────────────────────────────────────────────────────────────┐
│ burnrate │ Session: $0.42 / $2.00 max │ Burn: $0.08/min │ Cache: 92.1% │
│ Tokens: 18.4k total │ Turns: 6 │ Status: OPTIMAL (Prompt Cache Warm) │
└─────────────────────────────────────────────────────────────────────────────┘
```
---
## 🏛️ System Architecture
```
+-----------------------------------------------------------------------------------+
| Coding Agent CLI (Claude Code, Cursor, Codex, OpenCode) |
| `export ANTHROPIC_BASE_URL=http://localhost:7070` |
+-----------------------------------------------------------------------------------+
|
| Streaming HTTP Request
v
+-----------------------------------------------------------------------------------+
| `burnrate` Proxy (Rust Axum) |
| |
| [ Feature 1: Hard Budget Circuit Breaker ] |
| - If `session_cost >= max_spend` (e.g. $1.50): |
| * Terminate connection immediately with HTTP 402 Payment Required |
| * Emits diagnostic: "burnrate: Budget cap reached ($1.50). Session halted." |
| |
| [ Feature 2: Prompt-Cache Invalidation Linter ] |
| - Computes byte-level prefix hash diffs between Turn N-1 and Turn N: |
| * If early prefix bytes drift (e.g. timestamp injected at byte 0), |
| flashes warning: "⚠️ Cache busted at byte 18: prefix drift detected" |
| |
| [ Feature 3: In-Line Secret Guard & Token Redactor ] |
| - Scans outbound prompts for OpenAI, Anthropic, AWS, and GitHub credentials |
| * Masks leaked tokens with `[REDACTED_BY_BURNRATE:]` before forwarding |
| |
| [ Feature 4: Live Terminal HUD ] |
| - Real-time spend tracking ($/min burn, cache hit ratio, token breakdown) |
+-----------------------------------------------------------------------------------+
|
| Forward to Upstream
v
+-----------------------------------------------------------------------------------+
| Official Provider API (Anthropic, OpenAI, DeepSeek) |
+-----------------------------------------------------------------------------------+
```
---
## 🚀 Quickstart
### 1. Installation
```bash
cargo install burnrate-agent
# Or build from source:
git clone https://github.com/ronitgupta138/burnrate.git
cd burnrate && cargo build --release
```
### 2. Wrap Any Agent Command (Hard Cap Mode)
Wrap your agent with a strict budget limit (e.g., $2.00):
```bash
# Wrap Claude Code
burnrate run --max-spend 2.00 -- claude
# Wrap OpenCode or Codex
burnrate run --max-spend 1.50 -- opencode
```
### 3. Run Standalone Proxy Daemon
Start the proxy daemon in a dedicated tmux pane:
```bash
burnrate serve --port 7070 --max-spend 3.00
```
In your development terminal:
```bash
export ANTHROPIC_BASE_URL="http://127.0.0.1:7070"
export OPENAI_BASE_URL="http://127.0.0.1:7070"
```
---
## 📊 Supported Model Pricing Tables
`burnrate` maintains exact pricing tables factoring in 90% prompt-cache read discounts:
| Model Family | Input / MTok | Output / MTok | Cache Read (90% off) | Cache Write |
| :--- | :---: | :---: | :---: | :---: |
| **Claude 3.7 / 3.5 Sonnet** | $3.00 | $15.00 | **$0.30** | $3.75 |
| **Claude 3.5 Opus** | $15.00 | $75.00 | **$1.50** | $18.75 |
| **Claude 3.5 Haiku** | $0.80 | $4.00 | **$0.08** | $1.00 |
| **GPT-4o** | $2.50 | $10.00 | **$1.25** | $2.50 |
| **GPT-4o-mini** | $0.15 | $0.60 | **$0.075** | $0.15 |
| **DeepSeek-V3** | $0.14 | $0.28 | **$0.014** | $0.14 |
---
## 🧪 Verification & Benchmarks
```bash
cargo test
```
```
running 31 tests
test adapter::tests::test_openai_to_anthropic_response_translation ... ok
test adapter::tests::test_anthropic_to_openai_translation ... ok
test cache_engine::tests::test_cache_breakpoint_injection_on_large_system_prompt ... ok
test config::tests::test_config_defaults_and_serialization ... ok
test cache_engine::tests::test_cache_bypass_on_small_payload ... ok
test fallback::tests::test_fallback_trigger_at_95_percent_spend ... ok
test policy::tests::test_policy_engine_rule_enforcement ... ok
test vault::tests::test_token_vault_injection ... ok
test web::tests::test_embedded_dashboard_html_validity ... ok
test context_guard::tests::test_lockfile_truncation ... ok
test secret_guard::tests::test_secret_guard_unit ... ok
test ledger::tests::test_ledger_record_and_summary ... ok
test test_local_fallback_decision_logic ... ok
test test_schema_adapter_full_translation_loop ... ok
test test_proxy_fallback_header_when_spend_critical ... ok
test test_cache_engine_breakpoint_injection ... ok
test test_proxy_end_to_end_context_truncation_and_cache_injection ... ok
test test_context_window_guard_thresholds ... ok
test test_budget_governor_hard_spend_breaker ... ok
test test_pricing_calculation ... ok
test test_prompt_cache_invalidation_linter ... ok
test test_terminal_hud_rendering ... ok
test test_proxy_budget_enforcement_endpoint ... ok
test test_config_load_and_serialization ... ok
test test_ledger_record_and_summary_lifecycle ... ok
test test_proxy_integration_records_to_ledger ... ok
test test_mcp_initialize_and_tool_list ... ok
test test_mcp_call_status_tool ... ok
test test_prometheus_metrics_export ... ok
test test_model_cost_router_downgrade_on_low_budget ... ok
test test_proxy_secret_redaction_header_and_body ... ok
test test_secret_guard_all_credential_types ... ok
test test_streaming_token_counter_ingestion ... ok
test test_swarm_role_partitioning_and_summary ... ok
test test_proxy_multi_agent_swarm_header_enforcement ... ok
test test_policy_engine_custom_rules_and_verdict ... ok
test test_token_vault_key_injection ... ok
test test_proxy_v2_kernel_dashboard_and_policy_enforcement ... ok
test test_team_budget_pool_multi_agent_partitioning ... ok
test test_velocity_brake_threshold_enforcement ... ok
test result: ok. 31 passed; 0 failed; 0 ignored; finished in 0.01s
```
---
## 💖 Sponsoring & Supporting Development
`burnrate` is an independent open-source project built to protect developers, teams, and autonomous agent harnesses from accidental $100+ API bill explosions and silent prompt-cache invalidation leaks.
If `burnrate` has saved your team from runaway token invoices or improved your agent infrastructure, consider **[sponsoring development on GitHub](https://github.com/sponsors/ronitgupta138)**:
* **☕ Backers ($5–$10/mo):** Named credit in official release notes and README backer roll.
* **⚡ Power Users & Teams ($25/mo):** Direct influence on roadmap RFCs and priority feature triage.
* **🏢 Organization Sponsors ($100/mo):** Prominent company logo and link displayed on README and dashboard.
👉 **[Become a GitHub Sponsor](https://github.com/sponsors/ronitgupta138)**
---
## 📜 License
MIT License. Designed and engineered by [Ronit Gupta](https://github.com/ronitgupta138).