title	Aider + fak: governed CLI coding agent
description	Wire fak as a tool-governance layer for Aider, the CLI-based AI pair programmer, adding capability-floor enforcement and quarantine protection.

Aider + fak Integration Guide

This guide shows how to use fak as a tool-governance layer for Aider, the CLI-based AI pair programmer. Every command Aider proposes is evaluated by the kernel before it executes — dangerous calls are dropped, malformed calls are repaired, and policy violations are refused.

Overview

┌──────────────┐   OpenAI / Anthropic   ┌────────────────────────┐
│     Aider    │ ──────────────────────▶ │  fak serve (gateway)   │
│   (CLI)      │ ◀──── SSE stream ─────  │  adjudicates tools     │
└──────────────┘                        └────────────────────────┘
       ▲                                                 │
       │ OPENAI_API_BASE / ANTHROPIC_BASE_URL           │
       │ (points at fak)                                ▼
       │                                        ┌───────────────┐
       │                                        │  Local Model  │
       │                                        │ or Cloud API  │
       │                                        └───────────────┘

The gateway sits between Aider and the model:

Aider → fak: Aider sends API requests with proposed file edits and commands
fak kernel: Adjudicates each proposed call (allow, deny, transform, quarantine)
fak → model: Sends only the admitted (or repaired) calls to the model
fak → Aider: Returns results, with the kernel's decisions applied

Result: Aider can edit your codebase, but the kernel blocks destructive commands, prevents self-modification, and contains untrusted tool results.

Prerequisites

1. Install fak

# From the repo (the Go module is the repo root)
git clone https://github.com/anthony-chaudhary/fak && cd fak
go build -o fak ./cmd/fak

# Or via the installer
curl -fsSL https://raw.githubusercontent.com/anthony-chaudhary/fak/main/install.sh | sh

Verify installation:

./fak version

2. Install Aider

pip install aider-chat

Verify installation:

aider --version

3. Choose your upstream model

Aider can connect to fak in two modes:

Proxy mode: fak forwards to an external model (OpenAI, Anthropic, Ollama, vLLM, etc.)
In-kernel mode: fak serves its own fused model

For development, proxy mode is recommended as it gives you full model capabilities while still enforcing tool governance.

Quick Start: One-command setup

The fastest way to put the kernel in front of Aider:

# Terminal 1: Start the fak gateway
./fak serve \
  --addr 127.0.0.1:8080 \
  --provider openai \
  --base-url http://localhost:11434/v1 \
  --model qwen2.5:1.5b \
  --policy examples/customer-support-readonly-policy.json

# Terminal 2: Run Aider through fak
export OPENAI_API_BASE="http://127.0.0.1:8080/v1"
export OPENAI_API_KEY="fak-local"
aider --model openai/qwen2.5:1.5b

Aider now runs through the capability floor — every file edit, git operation, or shell command is adjudicated before execution.

Method 1: OpenAI-compatible provider (Recommended)

Aider's primary integration path is through OpenAI-compatible APIs. fak provides an OpenAI-compatible /v1/chat/completions endpoint.

Step 1: Start the fak HTTP gateway

./fak serve \
  --addr 127.0.0.1:8080 \
  --provider openai \
  --base-url http://localhost:11434/v1 \
  --model qwen2.5:1.5b \
  --policy examples/customer-support-readonly-policy.json

Verify health:

curl http://127.0.0.1:8080/healthz
# {"ok":true,"model":"qwen2.5:1.5b","engine":"inkernel"}

Step 2: Configure Aider

Via environment variables (Recommended):

export OPENAI_API_BASE="http://127.0.0.1:8080/v1"
export OPENAI_API_KEY="fak-local"
export OPENAI_MODEL="qwen2.5:1.5b"

Via configuration file:

In ~/.aider.conf.yml:

openai-api-base: http://127.0.0.1:8080/v1
model: openai/qwen2.5:1.5b
api-key: fak-local

Via CLI flags:

aider \
  --openai-api-base http://127.0.0.1:8080/v1 \
  --model openai/qwen2.5:1.5b \
  --api-key fak-local

Step 3: Run Aider

aider  # Starts Aider with your git repo and the governed model

Aider's file edits and commands now flow through fak:

Aider → fak /v1/chat/completions → adjudication → upstream model
                                            ↓
                                      capability floor
                                            ↓
                                    allowed/denied/transformed
                                            ↓
                                      Aider (with filtered results)

Method 2: govern a real Claude model upstream

To put the floor in front of an actual Anthropic Claude model, keep the upstream provider Anthropic — fak proxies to api.anthropic.com — but wire Aider to fak the same way as Method 1, over the OpenAI-compatible endpoint fak always exposes. Aider is a LiteLLM-based client that speaks the OpenAI wire; it has no Anthropic-specific flag, so the connection is identical to Method 1 regardless of what fak proxies upstream.

Step 1: Start the fak gateway with the Anthropic upstream

./fak serve \
  --addr 127.0.0.1:8080 \
  --provider anthropic \
  --base-url https://api.anthropic.com/v1 \
  --api-key-env ANTHROPIC_API_KEY \
  --model claude-sonnet-4-20250514 \
  --policy examples/customer-support-readonly-policy.json

Step 2: Configure Aider (OpenAI-compatible, same as Method 1)

export OPENAI_API_BASE="http://127.0.0.1:8080/v1"
export OPENAI_API_KEY="fak-local"

Step 3: Run Aider

Name the served model with the openai/ prefix so Aider routes it over the OpenAI wire to fak (which proxies to the Claude model upstream):

aider --model openai/claude-sonnet-4-20250514

Creating a Capability Floor for Aider

A capability floor defines which operations Aider may perform. Start from the built-in default:

# Dump the default policy as a starting point
./fak policy --dump > aider-policy.json

Example: Safe coding agent policy

{
  "version": "fak-policy/v1",
  "posture": "fail_closed",
  "allow": [
    "read_file",
    "write_file",
    "list_directory",
    "git_diff",
    "git_log",
    "git_show",
    "git_blame"
  ],
  "allow_prefix": [
    "read_",
    "get_",
    "search_",
    "list_",
    "git_",
    "lint_",
    "format_"
  ],
  "deny": {
    "run_command": "POLICY_BLOCK",
    "git_push": "POLICY_BLOCK",
    "git_reset": "POLICY_BLOCK",
    "git_clean": "POLICY_BLOCK",
    "delete_file": "POLICY_BLOCK"
  },
  "self_modify_globs": [
    ".git/",
    "aider-policy.json",
    ".env",
    "id_rsa"
  ],
  "arg_rules": [
    {
      "tool": "read_file",
      "arg": "path",
      "deny_regex": ".*\\.env$",
      "reason": "SECRET_EXFIL"
    },
    {
      "tool": "run_command",
      "arg": "command",
      "deny_regex": "rm\\s+-rf|sudo|git\\s+push",
      "reason": "POLICY_BLOCK"
    }
  ]
}

Validate before using:

./fak policy --check aider-policy.json

Common Patterns for Aider Workflows

Pattern 1: Read-only code review

Configure fak to allow reads but block writes and destructive Git operations:

{
  "allow": ["read_file", "list_directory", "git_diff", "git_log"],
  "deny": {
    "write_file": "POLICY_BLOCK",
    "run_command": "POLICY_BLOCK",
    "git_push": "POLICY_BLOCK"
  }
}

Pattern 2: Safe refactoring with Git awareness

Allow safe file operations but block destructive Git operations:

{
  "allow_prefix": ["read_", "write_", "git_diff", "git_log", "git_show", "git_blame"],
  "deny": {
    "git_push": "POLICY_BLOCK",
    "git_reset": "POLICY_BLOCK",
    "git_clean": "POLICY_BLOCK",
    "run_command": "POLICY_BLOCK"
  },
  "self_modify_globs": [".git/"]
}

Pattern 3: Quarantine for external tool results

Protect against poisoned responses from external APIs:

# Enable quarantine on the gateway
./fak serve --addr 127.0.0.1:8080 \
  --base-url https://api.openai.com/v1 \
  --policy policy.json \
  --vdso=true  # Enables content-addressed cache and quarantine

If an external tool returns suspicious content (e.g., injection attempts), fak automatically quarantines it, preventing it from entering Aider's context.

Monitoring and Debugging

Health checks

curl http://127.0.0.1:8080/healthz

Metrics

curl http://127.0.0.1:8080/metrics

Key metrics:

fak_gateway_time_to_ready_seconds - Startup time
fak_vdso_hit_rate - Cache hit rate
fak_gateway_operations_total{verdict="DENY"} - Denied calls (by reason label)
fak_kernel_quarantines_total - Quarantined results

Debugging a denied operation

When Aider reports a denied operation, reproduce it offline:

./fak preflight \
  --tool write_file \
  --args '{"path":"test.txt","content":"test"}' \
  --policy your-policy.json
# verdict=DENY reason=POLICY_BLOCK

Use --explain to get a detailed breakdown:

./fak preflight --explain \
  --tool run_command \
  --args '{"command":"rm -rf /tmp"}' \
  --policy your-policy.json

Troubleshooting

Aider can't connect to the gateway

Verify fak is running:
```
curl http://127.0.0.1:8080/healthz
```

Check your environment variables:

echo $OPENAI_API_BASE
# Should output: http://127.0.0.1:8080/v1

Verify the policy file exists:
```
./fak policy --check your-policy.json
```

All operations are being denied

Check the policy's posture:
- posture: "fail_closed" (default) denies everything not explicitly allowed
- Ensure your tools are in allow or match an allow_prefix

Test a specific call:

./fak preflight --tool read_file --args '{"path":"test.txt"}' --policy your-policy.json

Slow first response

Expected on large local models — the Aider prompt is ~25K tokens. Subsequent requests are faster if you enable --vdso=true (content-addressed caching).

Advanced Usage

Authentication

For production use, require an API key:

./fak serve \
  --addr 0.0.0.0:8080 \
  --base-url ... \
  --model ... \
  --require-key-env FAK_TOKEN

Aider sends x-api-key: (OpenAI SDK), which fak honors.

Cloud providers

# OpenAI
./fak serve \
  --provider openai \
  --base-url https://api.openai.com/v1 \
  --api-key-env OPENAI_API_KEY \
  --model gpt-4

# Anthropic
./fak serve \
  --provider anthropic \
  --base-url https://api.anthropic.com/v1 \
  --api-key-env ANTHROPIC_API_KEY \
  --model claude-sonnet-4-20250514

Local models

# Ollama
./fak serve \
  --provider openai \
  --base-url http://localhost:11434/v1 \
  --model qwen2.5-coder:7b

# vLLM
./fak serve \
  --provider openai \
  --base-url http://localhost:8000/v1 \
  --model qwen2.5-coder:7b

Cross-references

Integration index: README.md — universal recipe and which-agent routing
Compatibility matrix: compatibility-matrix.md — full field survey
Policy schema: ../../POLICY.md — authoring capability floors
Aider docs: https://aider.chat — official Aider documentation
** fak architecture**: ../../ARCHITECTURE.md — kernel internals

License

Apache-2.0

FilesExpand file tree

aider.md

Latest commit

History

aider.md

File metadata and controls

Aider + fak Integration Guide

Overview

Prerequisites

1. Install fak

2. Install Aider

3. Choose your upstream model

Quick Start: One-command setup

Method 1: OpenAI-compatible provider (Recommended)

Step 1: Start the fak HTTP gateway

Step 2: Configure Aider

Step 3: Run Aider

Method 2: govern a real Claude model upstream

Step 1: Start the fak gateway with the Anthropic upstream

Step 2: Configure Aider (OpenAI-compatible, same as Method 1)

Step 3: Run Aider

Creating a Capability Floor for Aider

Example: Safe coding agent policy

Common Patterns for Aider Workflows

Pattern 1: Read-only code review

Pattern 2: Safe refactoring with Git awareness

Pattern 3: Quarantine for external tool results

Monitoring and Debugging

Health checks

Metrics

Debugging a denied operation

Troubleshooting

Aider can't connect to the gateway

All operations are being denied

Slow first response

Advanced Usage

Authentication

Cloud providers

Local models

Cross-references

License