reposwarm-cli

CLI for RepoSwarm — AI-powered multi-repo architecture discovery.

Written in Go. Single 9MB binary, zero runtime dependencies, 4ms startup.

Install

curl -fsSL https://raw.githubusercontent.com/reposwarm/reposwarm-cli/main/install.sh | sh

Or build from source:

git clone https://github.com/reposwarm/reposwarm-cli.git
cd reposwarm-cli
go build -o reposwarm ./cmd/reposwarm

🤖 Agent-Native

Every command supports --for-agent for machine-friendly output. No spinners, no colors, no interactive prompts — just clean text that AI agents can parse.

# Agent gets JSON output
reposwarm status --for-agent --json

# Full local setup — zero human interaction needed
reposwarm new --local --for-agent

# Configure provider non-interactively
reposwarm config provider setup --provider bedrock --auth-method iam-role --region us-east-1 --non-interactive

# Debug logs (truncated at 5000 chars, with path to full file)
reposwarm debug-logs --for-agent

# Investigate and get results
reposwarm investigate https://github.com/org/repo --for-agent
reposwarm results read my-repo --for-agent

Designed for OpenClaw, Claude Code, Codex, and any AI coding agent.

Prerequisites

For local setup (reposwarm new --local):

Tool	Version	Notes
Docker	24+	Must be running (not just installed)
Docker Compose	v2	Usually bundled with Docker Desktop
Git	2.x	For cloning repos

⚠️ Docker must be running. The CLI checks this during setup and will warn you if Docker is installed but the daemon isn't started.

No Node.js or Python needed — all services run as pre-built Docker containers.

For remote setup (connecting to an existing server): just the CLI binary. No other dependencies.

Cookbook

Real workflows, copy-paste ready. Each recipe is a complete flow.

🚀 First Time Setup (Local)

reposwarm new --local             # Bootstrap everything: Temporal, API, Worker, UI
reposwarm status                  # Verify connection
reposwarm doctor                  # Full health check
reposwarm show ui                 # Open the web UI

🚀 First Time Setup (Remote Server)

reposwarm config init             # Interactive setup (enter API URL + token)
reposwarm status                  # Verify connection

🔍 Run Your First Investigation

reposwarm repos add my-app --url https://github.com/org/my-app
reposwarm investigate my-app      # Pre-flight checks run automatically
reposwarm wf progress             # Watch it work
reposwarm results sections my-app # Browse the results
reposwarm results read my-app     # Read the full report

🔍 Investigate All Repos

reposwarm investigate --all --parallel 3
reposwarm dashboard               # Live TUI — progress for every repo

🔍 Check Before You Start (Dry Run)

reposwarm preflight my-app        # System + repo readiness check
reposwarm investigate my-app --dry-run  # Pre-flight only, no workflow created

🩺 Something's Stuck — Full Diagnosis

The 2-minute flow that finds any issue:

# Step 1: What's wrong?
reposwarm doctor

# Step 2: Full install log (paged for humans, text dump for agents)
reposwarm debug-logs

# Step 3: Check for stalled workflows + worker errors
reposwarm errors

# Step 3: Zoom in on a specific workflow
reposwarm wf status <workflow-id> -v

# Step 4: See the Temporal event timeline
reposwarm wf history <workflow-id>

# Step 5: Check worker logs directly
reposwarm logs worker -f

What doctor catches:

Missing env vars — driven by providers.json (single source of truth for LLM + git provider requirements)
Git provider auth: GITHUB_TOKEN, GITLAB_TOKEN, AZURE_DEVOPS_PAT, etc. (based on configured git provider)
Worker log runtime errors (crashes, tracebacks — env validation noise filtered out)
Per-worker health in multi-worker setups
Stalled workflows with zero progress
Config, API, Temporal, DynamoDB connectivity
Version compatibility (API ↔ CLI)
CLI update availability (checks GitHub releases)

What errors catches:

Activity failures, timeouts, workflow failures
Stalled activities (scheduled but never started — worker can't pick them up)
Zero-progress workflows (running > 30 min with nothing completed)
Worker startup/validation failures

🔧 Fix Missing Environment Variables

When doctor tells you env vars are missing:

# See what's set and what's missing
reposwarm config worker-env list

# Set the missing vars
reposwarm config worker-env set ANTHROPIC_API_KEY sk-ant-your-key-here
reposwarm config worker-env set GITHUB_TOKEN ghp_your-token-here

# Restart the worker to pick up changes
reposwarm restart worker --wait

# Verify it's healthy now
reposwarm doctor

🔧 Configure LLM Provider

RepoSwarm supports three provider backends:

Provider	Auth	Best for
`anthropic`	API key	Direct Anthropic API access
`bedrock`	AWS credentials (IAM role, access keys, SSO, profile, or Bedrock API keys)	AWS-native environments
`litellm`	Proxy URL + optional key	OpenAI, Gemini, Mistral, and 100+ other models via LiteLLM proxy

OpenAI and Gemini are not built-in providers — use litellm to route to them. See the LiteLLM section below.

Anthropic (Direct API)

# Interactive
reposwarm config provider setup

# Non-interactive
reposwarm config provider setup \
  --provider anthropic \
  --api-key sk-ant-api03-YOUR-KEY \
  --model sonnet \
  --non-interactive

Amazon Bedrock

# IAM role (EC2 instance profile / ECS task role — recommended)
reposwarm config provider setup \
  --provider bedrock \
  --auth-method iam-role \
  --region us-east-1 \
  --model opus \
  --non-interactive

# AWS access keys
reposwarm config provider setup \
  --provider bedrock \
  --auth-method access-keys \
  --aws-key AKIA... \
  --aws-secret wJalr... \
  --region us-east-1 \
  --model sonnet \
  --non-interactive

# AWS SSO / Identity Center
reposwarm config provider setup \
  --provider bedrock \
  --auth-method sso \
  --aws-profile my-sso-profile \
  --region us-west-2 \
  --model sonnet \
  --non-interactive

# Named AWS profile (~/.aws/credentials)
reposwarm config provider setup \
  --provider bedrock \
  --auth-method profile \
  --aws-profile prod \
  --region us-east-1 \
  --model opus \
  --non-interactive

# Bedrock API keys (from AWS Console → Bedrock → API keys)
reposwarm config provider setup \
  --provider bedrock \
  --auth-method api-keys \
  --bedrock-key br-... \
  --region us-east-1 \
  --model sonnet \
  --non-interactive

How IAM role auth works in Docker: The worker container runs with network_mode: host, sharing the host's network stack. This lets the AWS SDK inside the container transparently access EC2 instance metadata (IMDSv2) or any credentials configured on the host (~/.aws/), exactly like it would in ECS Fargate. No credential injection or special configuration needed.

Using OpenAI, Gemini, or Other Models via LiteLLM

RepoSwarm uses Claude under the hood, so non-Anthropic models require a LiteLLM proxy to translate between APIs. LiteLLM supports 100+ models including OpenAI, Gemini, Mistral, Cohere, and more.

1. Start a LiteLLM proxy (one-time setup):

pip install litellm

# OpenAI
litellm --model openai/gpt-4o --port 4000

# Google Gemini
GEMINI_API_KEY=your-key litellm --model gemini/gemini-2.5-pro --port 4000

# Or use a config file for multiple models:
# See https://docs.litellm.ai/docs/proxy/configs
litellm --config litellm_config.yaml --port 4000

2. Point RepoSwarm at the proxy:

# OpenAI via LiteLLM
reposwarm config provider setup \
  --provider litellm \
  --proxy-url http://localhost:4000 \
  --proxy-key sk-your-litellm-key \
  --model openai/gpt-4o \
  --non-interactive

# Google Gemini via LiteLLM
reposwarm config provider setup \
  --provider litellm \
  --proxy-url http://localhost:4000 \
  --model gemini/gemini-2.5-pro \
  --non-interactive

# Any other LiteLLM-supported model
reposwarm config provider setup \
  --provider litellm \
  --proxy-url https://my-proxy.example.com \
  --proxy-key sk-proxy-key \
  --model mistral/mistral-large-latest \
  --non-interactive

Note: Investigation quality may vary with non-Claude models. RepoSwarm's prompts and workflows are optimized for Claude Sonnet/Opus.

Provider Management

# Quick-switch provider (keeps other settings)
reposwarm config provider set bedrock
reposwarm config provider set anthropic --check

# Check what's configured (detects CLI vs worker drift)
reposwarm config provider show

# List available model aliases
reposwarm config model list

# Pin model versions (prevents silent version drift)
reposwarm config model pin

# Restart worker to apply changes
reposwarm restart worker

🔧 Configure Git Provider

Tell RepoSwarm which git hosting platform your repos are on, so it knows what credentials to require:

# Interactive setup
reposwarm config git setup

# Direct set
reposwarm config git set github        # GitHub (needs GITHUB_TOKEN)
reposwarm config git set codecommit    # AWS CodeCommit (needs IAM + AWS CLI)
reposwarm config git set gitlab        # GitLab (needs GITLAB_TOKEN)
reposwarm config git set azure         # Azure DevOps (needs AZURE_DEVOPS_PAT + ORG)
reposwarm config git set bitbucket     # Bitbucket (needs USERNAME + APP_PASSWORD)

# Show current config + required env vars
reposwarm config git show

After setting the git provider, doctor will check for the correct authentication variables:

reposwarm doctor
# Worker Environment
#   ✗ GITHUB_TOKEN — NOT SET — GitHub personal access token
#      Set it: reposwarm config worker-env set GITHUB_TOKEN <value>

📋 View Changelog

reposwarm changelog                    # Current version's release notes
reposwarm changelog v1.3.50            # Specific version
reposwarm changelog latest             # Latest release
reposwarm changelog --all              # All releases with dates
reposwarm changelog --since v1.3.45    # Cumulative changes since a version

⬆️ Upgrade Components

reposwarm upgrade                      # Upgrade CLI to latest
reposwarm upgrade api                  # Upgrade API server (git pull + build + restart)
reposwarm upgrade ui                   # Upgrade UI
reposwarm upgrade all                  # Upgrade all components

🌐 Remote Access via SSH Tunnel

When RepoSwarm runs on a remote server (EC2, VPS), access the web UI through an SSH tunnel:

reposwarm tunnel                       # Shows copy-paste SSH tunnel command

🗑️ Uninstall

reposwarm uninstall                    # Interactive removal
reposwarm uninstall --force            # Skip confirmation
reposwarm uninstall --keep-config      # Keep ~/.reposwarm for reinstall

For local installs: stops all services, runs docker compose down, removes files. For remote CLI-only setups: removes the binary and config, warns about the server.

🔧 Fix a Stuck Investigation (End-to-End)

The full fix loop — from broken to running in 6 commands:

reposwarm doctor                  # → finds: ANTHROPIC_API_KEY NOT SET
reposwarm config worker-env set ANTHROPIC_API_KEY sk-ant-xxx
reposwarm config worker-env set GITHUB_TOKEN ghp_xxx
reposwarm restart worker --wait   # → ✅ worker-1 healthy
reposwarm wf retry <workflow-id>  # → terminates old, starts new
reposwarm wf progress             # → watch it go

🔧 Replace a Stuck Investigation

# Terminates any existing workflow for this repo, then starts fresh
reposwarm investigate my-app --replace

🔧 Retry a Failed Investigation with a Different Model

reposwarm wf retry <workflow-id> --model us.anthropic.claude-opus-4-6

📊 Monitor Running Investigations

# Live dashboard (htop-style, all repos)
reposwarm dashboard

# Focus on one repo
reposwarm dashboard --repo my-app

# Watch a specific workflow with step checklist
reposwarm wf progress --repo my-app --wait

# Quick snapshot for scripts/agents
reposwarm dashboard --json
reposwarm errors --json

📊 Check Worker Health

# List all workers
reposwarm workers list

# Deep-dive on a specific worker
reposwarm workers show worker-1

# See all services (Docker containers or processes)
reposwarm services

🔎 Search and Explore Results

# Search across all investigations
reposwarm results search "authentication" --max 10

# Read a specific section
reposwarm results read my-app authentication

# Compare two repos
reposwarm results diff repo-a repo-b

# Export everything to markdown files
reposwarm results export --all -d ./docs

# Audit completeness (all 17 sections present?)
reposwarm results audit

📖 Query Architecture Results — `ask` CLI

Architecture querying has moved to the standalone ask CLI.

RepoSwarm writes architecture docs (investigations). ask reads them.

# Install ask
curl -fsSL https://raw.githubusercontent.com/reposwarm/ask-cli/main/install.sh | sh

# Set up local askbox (auto-detects RepoSwarm config)
ask setup

# Query your architecture
ask "how does auth work across services?"
ask results list
ask results search "DynamoDB"
ask results export --all -d ./docs

See the ask README for full documentation.

Note: reposwarm ask --arch and reposwarm results still work but will show a deprecation notice pointing to the ask CLI.

🧹 Cleanup and Maintenance

# Clean up old workflows (completed/failed/terminated)
reposwarm wf prune --older 7d --dry-run    # Preview first
reposwarm wf prune --older 7d -y           # Actually prune

# Prune only failed workflows
reposwarm wf prune --status failed -y

# Gracefully cancel a running investigation (finishes current step)
reposwarm wf cancel <workflow-id>

# Hard terminate (immediate)
reposwarm wf terminate <workflow-id> -y

⚙️ Service Management (Local Mode)

# See what's running
reposwarm services

# Restart everything
reposwarm restart

# Restart just the worker
reposwarm restart worker --wait

# Stop/start individual services
reposwarm stop worker
reposwarm start worker --wait

# Check all service URLs
reposwarm url all

🤖 Agent / Script Integration

Every command supports --json for structured output and --for-agent for plain text:

# Check system health programmatically
reposwarm doctor --json | jq '.checks[] | select(.status=="fail")'

# Get all errors as JSON
reposwarm errors --json | jq '.workerErrors, .stalls, .errors'

# Single dashboard snapshot (no TUI, exits immediately)
reposwarm dashboard --json

# Pre-flight check before scripted investigation
reposwarm preflight my-app --json | jq '.ready'

# Investigate with no prompts
reposwarm investigate my-app --force --json

# Worker status for monitoring
reposwarm workers list --json | jq '.workers[] | {name, status, envStatus}'

# Service health for monitoring
reposwarm services --json | jq '.[] | {name, status, pid}'

Command Reference

Setup & Configuration

Command	Description
`reposwarm config init`	Interactive setup wizard
`reposwarm config show`	Display config (includes server config + model drift warning)
`reposwarm config set <key> <value>`	Update a config value
`reposwarm config server`	View server-side config
`reposwarm config server-set <key> <value>`	Update server config
`reposwarm config worker-env list`	Show worker env vars (`--reveal` for unmasked values)
`reposwarm config worker-env set <K> <V>`	Set worker env var (`--restart` to auto-restart)
`reposwarm config worker-env unset <K>`	Remove worker env var
`reposwarm config provider setup`	Interactive provider setup (Anthropic, Bedrock, LiteLLM)
`reposwarm config provider set <provider>`	Quick-switch provider (`anthropic`, `bedrock`, `litellm`)
`reposwarm config provider show`	Show provider config + drift detection
`reposwarm config model set <alias\|id>`	Set model (resolves aliases per provider, `--sync` to worker)
`reposwarm config model show`	Show model across CLI, server, and worker
`reposwarm config model list`	List aliases with resolved IDs per provider
`reposwarm config model pin`	Pin all model aliases to current versions
`reposwarm new --local`	Bootstrap complete local installation
`reposwarm upgrade`	Self-update (`--force` to reinstall)

Diagnostics

Command	Description
`reposwarm status`	Quick API health + latency
`reposwarm doctor`	Full diagnosis: config, API, Temporal, workers, env, logs, stalls
`reposwarm preflight [repo]`	Verify system readiness for an investigation
`reposwarm errors`	Errors + stalls + worker failures (`--repo`, `--stall-threshold`)
`reposwarm logs [service]`	View service logs (`-f` follow, `-n` lines)

Workers & Services

Command	Description
`reposwarm workers list`	All workers: health, queue, env status (`-v` for PID/host)
`reposwarm workers show <name>`	Worker deep-dive: env, logs, tasks
`reposwarm services`	Process table: PID, status, port, manager
`reposwarm restart [service]`	Restart one or all (`--wait`, `--timeout`)
`reposwarm stop <service>`	Graceful stop
`reposwarm start <service>`	Start service (`--wait` for health check)

Repositories

Command	Description
`reposwarm repos list`	List repos (`--source`, `--filter`, `--enabled`)
`reposwarm repos show <name>`	Detailed repo view
`reposwarm repos add <name>`	Add repo (`--url`, `--source`)
`reposwarm repos remove <name>`	Remove (`-y` skip confirm)
`reposwarm repos enable/disable <name>`	Toggle investigation eligibility
`reposwarm repos discover`	Auto-discover repos from your configured git provider (GitHub, GitLab, CodeCommit, Azure DevOps, Bitbucket)

Investigation & Workflows

Command	Description
`reposwarm investigate <repo>`	Start investigation (pre-flight auto-runs)
	`--force` skip pre-flight, `--replace` terminate existing, `--dry-run`
`reposwarm investigate --all`	All enabled repos (`--parallel`)
`reposwarm wf list`	List recent workflows (`--limit`)
`reposwarm wf status <id>`	Workflow details (`-v` for activities + worker attribution)
`reposwarm wf history <id>`	Temporal event timeline (`--filter`, `--limit`)
`reposwarm wf progress`	Progress across repos (`--repo`, `--wait`)
`reposwarm wf watch [id]`	Live watch (`--interval`)
`reposwarm wf retry <id>`	Terminate + re-investigate (`-y`, `--model`)
`reposwarm wf cancel <id>`	Graceful cancel (current activity completes)
`reposwarm wf terminate <id>`	Hard stop (`-y`, `--reason`)
`reposwarm wf prune`	Cleanup old workflows (`--older`, `--status`, `--dry-run`)
`reposwarm dashboard`	Live TUI (`--repo` focused, `--json` single snapshot)

Results & Analysis

Command	Description
`reposwarm results list`	Repos with results
`reposwarm results sections <repo>`	Section list
`reposwarm results read <repo> [section]`	Read results (`--raw` for markdown)
`reposwarm results search <query>`	Full-text search (`--repo`, `--section`, `--max`)
`reposwarm results export <repo> -o file.md`	Export to file
`reposwarm results export --all -d ./docs`	Export all
`reposwarm results audit`	Validate completeness
`reposwarm results diff <repo1> <repo2>`	Compare investigations
`reposwarm results report [repos...] -o f.md`	Consolidated report

Architecture Queries (Askbox)

Command	Description
`reposwarm ask <question>`	Quick Q&A about RepoSwarm usage
`reposwarm ask --arch <question>`	Architecture analysis via askbox agent
	`--repos` scope to repos, `--adapter` choose agent, `--no-wait` async

Prompts

Command	Description
`reposwarm prompts list`	List prompts
`reposwarm prompts show <name>`	View template (`--raw`)
`reposwarm prompts create/update/delete <name>`	Manage prompts
`reposwarm prompts toggle <name>`	Enable/disable

Global Flags

Flag	Description
`--json`	JSON output
`--for-agent`	Plain text (no colors/formatting)
`--api-url <url>`	Override API URL
`--api-token <token>`	Override API token
`--verbose`	Debug info
`-v` / `--version`	Print version

Environment Variables

Variable	Description
`REPOSWARM_API_URL`	Override API URL
`REPOSWARM_API_TOKEN`	Override bearer token

Provider Environment Variables

These are set automatically by config provider setup — you normally don't need to touch them:

Variable	When	Description
`CLAUDE_CODE_USE_BEDROCK`	Bedrock	Set to `1` to enable Bedrock mode
`AWS_REGION`	Bedrock	Required (Bedrock doesn't read `.aws/config`)
`ANTHROPIC_MODEL`	All	Primary model ID
`ANTHROPIC_SMALL_FAST_MODEL`	All	Fast/cheap model for triage
`ANTHROPIC_BASE_URL`	LiteLLM	Proxy endpoint URL
`ANTHROPIC_DEFAULT_OPUS_MODEL`	Bedrock	Pinned Opus version
`ANTHROPIC_DEFAULT_SONNET_MODEL`	Bedrock	Pinned Sonnet version
`ANTHROPIC_DEFAULT_HAIKU_MODEL`	Bedrock	Pinned Haiku version

Git Provider Environment Variables

Set via config git set <provider> — determines which credentials are required:

Variable	Git Provider	Description
`GITHUB_TOKEN`	GitHub	Personal access token (scopes: repo)
`GITLAB_TOKEN`	GitLab	Personal/project access token
`GITLAB_URL`	GitLab	Instance URL (optional, default: gitlab.com)
`AZURE_DEVOPS_PAT`	Azure DevOps	Personal access token
`AZURE_DEVOPS_ORG`	Azure DevOps	Organization URL
`BITBUCKET_USERNAME`	Bitbucket	Bitbucket username
`BITBUCKET_APP_PASSWORD`	Bitbucket	App password
`AWS_REGION`	CodeCommit	AWS region (uses IAM for auth)

Model Alias Resolution

Alias	Anthropic API	Amazon Bedrock
`sonnet`	`claude-sonnet-4-6`	`us.anthropic.claude-sonnet-4-6`
`opus`	`claude-opus-4-6`	`us.anthropic.claude-opus-4-6-v1`
`haiku`	`claude-haiku-4-5`	`us.anthropic.claude-haiku-4-5-20251001-v1:0`

Configurable Keys

Set via reposwarm config set <key> <value>:

Key	Default	Description
`apiUrl`	`http://localhost:3000/v1`	API server URL
`apiToken`	—	Bearer token
`region`	`us-east-1`	AWS region
`defaultModel`	`us.anthropic.claude-sonnet-4-6`	Default LLM model
`chunkSize`	`10`	Files per investigation chunk
`installDir`	`~/.reposwarm`	Local installation directory
`temporalPort`	`7233`	Temporal gRPC port
`temporalUiPort`	`8233`	Temporal UI port
`apiPort`	`3000`	API server port
`uiPort`	`3001`	Web UI port
`hubUrl`	—	Project hub URL

Development

go test ./...              # Tests
go vet ./...               # Lint
go build ./cmd/reposwarm   # Build

CI: GitHub push → Go 1.24 tests → Cross-compile (darwin/linux × arm64/amd64) → GitHub Release

Ecosystem

Project	Docker Image
reposwarm (worker)	`ghcr.io/reposwarm/worker:latest`
reposwarm-api	`ghcr.io/reposwarm/api:latest`
reposwarm-ui	`ghcr.io/reposwarm/ui:latest`
reposwarm-cli (this repo)	— (binary install)
reposwarm-askbox	`ghcr.io/reposwarm/askbox:latest`

All Docker images are multi-arch (linux/amd64 + linux/arm64), published automatically on every push to main.

License

MIT

Name		Name	Last commit message	Last commit date
Latest commit History 234 Commits
.github/workflows		.github/workflows
cmd/reposwarm		cmd/reposwarm
cross-agent-tests		cross-agent-tests
docs		docs
e2e		e2e
internal		internal
reports		reports
test-scenarios		test-scenarios
.gitignore		.gitignore
CLAUDE.md		CLAUDE.md
CLI-REFACTOR-PLAN.md		CLI-REFACTOR-PLAN.md
CODEOWNERS		CODEOWNERS
E2E-TEST-SCENARIO.md		E2E-TEST-SCENARIO.md
LICENSE		LICENSE
README.md		README.md
TASK.md		TASK.md
bugs-report.md		bugs-report.md
buildspec.yml		buildspec.yml
cli-roadmap.md		cli-roadmap.md
e2e-test.sh		e2e-test.sh
go.mod		go.mod
go.sum		go.sum
install.sh		install.sh

Folders and files

Latest commit

History

Repository files navigation

reposwarm-cli

Install

🤖 Agent-Native

Prerequisites

Cookbook

🚀 First Time Setup (Local)

🚀 First Time Setup (Remote Server)

🔍 Run Your First Investigation

🔍 Investigate All Repos

🔍 Check Before You Start (Dry Run)

🩺 Something's Stuck — Full Diagnosis

🔧 Fix Missing Environment Variables

🔧 Configure LLM Provider

Anthropic (Direct API)

Amazon Bedrock

Using OpenAI, Gemini, or Other Models via LiteLLM

Provider Management

🔧 Configure Git Provider

📋 View Changelog

⬆️ Upgrade Components

🌐 Remote Access via SSH Tunnel

🗑️ Uninstall

🔧 Fix a Stuck Investigation (End-to-End)

🔧 Replace a Stuck Investigation

🔧 Retry a Failed Investigation with a Different Model

📊 Monitor Running Investigations

📊 Check Worker Health

🔎 Search and Explore Results

📖 Query Architecture Results — ask CLI

🧹 Cleanup and Maintenance

⚙️ Service Management (Local Mode)

🤖 Agent / Script Integration

Command Reference

Setup & Configuration

Diagnostics

Workers & Services

Repositories

Investigation & Workflows

Results & Analysis

Architecture Queries (Askbox)

Prompts

Global Flags

Environment Variables

Provider Environment Variables

Git Provider Environment Variables

Model Alias Resolution

Configurable Keys

Development

Ecosystem

License

About

Resources

License

Uh oh!

Stars

Watchers

Forks

Releases 173

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

📖 Query Architecture Results — `ask` CLI

Packages