frontier-scout 1.2.1

A local AI adoption radar for tools, MCP servers, agent frameworks, and model drops.

Frontier Scout

Deep Scout — know about new AI tools, MCP servers, models, and risky dependency upgrades before everyone else. Personalised, local-first, try-before-trust.

Quickstart  ·  Demo  ·  Roadmap  ·  Bug report  ·  Feature request  ·  Releases

📑 Table of contents

• Why Frontier Scout
• Built with
• Quickstart
• 60-second demo
• Usage — killer workflow
• Safety model
• Cost
• Roadmap
• Contributing
• License
• Acknowledgments

---

🔭 Why Frontier Scout

Deep Scout — know about new AI tools, MCP servers, models, and risky dependency upgrades before everyone else. Frontier Scout reads your repo locally (filenames + AST imports, never source content) and turns the firehose of public AI releases into a personalised adoption radar with ADOPT / TRIAL / ASSESS / HOLD verdicts.

Three promises that anchor the product:

• Try before trust. Every adoption candidate gets a sandbox dry-run receipt, a permission map, and a guard check before it touches your real repo.
• Fix vulnerabilities you didn't know existed. Dependency intelligence cross-references your manifests against curated feeds — security, hardening, and breaking upgrades — and emits a trial recipe, not a lockfile rewrite.
• Bound risky engineering changes. Incident Change Scout turns an incident ticket into cited context, a bounded remediation plan, and a HITL approval interrupt before any write.

The TUI is the front door. Inside any repo:

frontier-scout

That lands you on Mission Control — two tabs only: Scout (the product) and Settings. The Scout tab shows one unified list of findings (AI tools + dependency upgrades), each row with full reasoning: what it is, why we suggest it, why it fits your repo, risk reasoning, and the next safe step. Press Enter to write a dry-run receipt. Run frontier-scout setup from anywhere to configure your LLM backend or schedule recurring scouts.

Every other CLI command (evaluate, trial, guard, report, packs, deps, incident, dossier) still works for scripting and CI; the TUI no longer tries to surface them all on one screen.

The posture is deliberately boring in the good way: CLI first, SQLite/local files by default, static reports, no hosted telemetry, no hidden auto-installs, and explicit approval before risky actions.

Why not just use newsletters or GitHub Trending?

Option	What it gives you	What is missing
Newsletters	Good awareness	Not repo-aware, not source-verifiable, rarely actionable.
GitHub Trending	Popularity signal	No risk/fit/adoption-cost judgment.
Manual research	Highest nuance	Slow, inconsistent, easy to skip when busy.
Frontier Scout	Source-backed verdicts and lab next steps	Requires your API key for live scans.

---

🧰 Built with

---

⚡ Quickstart

Prerequisites: Python 3.11+.

Install from PyPI with pipx (recommended) or pip:

pipx install frontier-scout
# or, no install:
uvx frontier-scout demo
# or, plain pip:
pip install frontier-scout

Configure once (LLM backend, automation vs ad-hoc):

frontier-scout setup

Then, inside any repo, open Mission Control:

frontier-scout

Mission Control lands on the Scout tab — the radar that ranks the latest AI releases that fit your repo. Tab keys 1–2 switch between Scout and Settings. From the highlighted verdict row, every core CLI capability is one keystroke: Enter for a dry-run trial, L for a dry-run lab (press again within 3s to spend on a live hermetic install), e for the Adoption-Firewall evaluation, D for a dossier saved under ~/.frontier-scout/dossiers/. Run it outside a repo and the picker offers 🌐 Universal scout (no repo) so you still get the latest releases on a golden plate, just not tailored. The setup is remembered — once you've run the wizard, frontier-scout from any directory drops you straight into Mission Control; use frontier-scout setup or Settings → Open setup wizard to reconfigure. The verdict detail panel surfaces explicit Concerns — burns tokens, abandoned, vendor lock-in, security surface, marketing-only, unproven — so you always see why we'd push back on adoption. The import-evidence scanner reads ASTs locally, provider availability shows up as cards, and nothing reads secrets, logs into services, installs tools, or sends repo content to an LLM. Limited terminals can use frontier-scout setup --plain; automation can use frontier-scout setup --json. The layout reflows for VS Code-style 80×24 panels.

On first launch, a one-time welcome overlay appears after the brand splash and walks you through the three core actions — Scout verdicts, trialling a tool, and running guard in CI. Press any key to enter Mission Control; the overlay never shows again.

Develop locally

git clone https://github.com/ajaysurya1221/frontier-scout
cd frontier-scout
python3 -m venv .venv && source .venv/bin/activate
pip install -e ".[dev]"
frontier-scout --help

---

⏱ 60-second demo

No API key. No Slack workspace. No cloud setup.

make demo
open .scratch/incident-demo/answer.md

The incident demo writes:

• .scratch/incident-demo/answer.md — cited remediation answer.
• .scratch/incident-demo/trace.jsonl — local OpenTelemetry-shaped spans.
• .scratch/incident-demo/audit.jsonl — Cloudflare-style audit records.
• .scratch/incident-demo/eval.json — golden eval score.

Then run the AI tool radar demo:

frontier-scout demo

This spins up a local HTTP server, opens your browser automatically at http://localhost:<port>/, and prints a guided next-steps panel in the terminal. Press Ctrl+C to stop serving.

The terminal panel looks like:

╭── ◉ FRONTIER · SCOUT  demo ready ───────────────────────────────────╮
│  Serving at  http://localhost:54321  ·  Ctrl+C to stop               │
│                                                                       │
│  ✓  briefing.html   adoption receipts                                 │
│  ✓  verdicts.json   raw verdict data                                  │
│  ✓  judge-trace.md  quality trace                                     │
│                                                                       │
│  Next steps:                                                          │
│    http://localhost:54321          ← browser opened · adoption cards  │
│    frontier-scout setup            ← Mission Control TUI              │
│    frontier-scout scan --dry-run   ← verdicts for this repo           │
│    ANTHROPIC_API_KEY=<key> ...     ← live scan                        │
╰───────────────────────────────────────────────────────────────────────╯

To write files without starting a server (CI or offline use):

frontier-scout demo --no-serve

The radar demo writes demo/briefing.html, demo/briefing.md, demo/verdicts.json, demo/cost-breakdown.md, and demo/judge-trace.md.

---

🛰 Usage — killer workflow

Someone drops a GitHub repo, MCP server, plugin, model, or agent framework in a newsletter or team chat. Frontier Scout turns that link into a local adoption decision instead of a vibes-based "looks safe" answer:

frontier-scout init --repo .
frontier-scout evaluate <tool-url>
frontier-scout trial <tool-or-url> --dry-run
frontier-scout guard --repo .
frontier-scout report

• init writes a local stack profile under ~/.frontier-scout (languages, package managers, container files, agent configs, and v0.4 import evidence from a tree-sitter pass).
• evaluate records source-backed local evidence and a permission manifest for one URL — capability map included.
• trial --dry-run writes an adoption receipt without installing anything; full trials use the hermetic lab.
• guard checks the local evidence ledger for risky tools that still need a stored trial receipt; CI-friendly exit codes.
• report renders the static HTML executive radar.

Inspect living packs and repo-relevant dependency upgrades:

frontier-scout packs list
frontier-scout packs show mcp
frontier-scout profile --repo . --dependencies
frontier-scout deps scan --repo .

---

🔒 Safety model

Frontier Scout handles untrusted public content and can optionally execute untrusted packages in the lab, so the safety rails are load-bearing:

• Source text is treated as untrusted data, not instructions.
• Tool names are checked against the source pool to reduce hallucinated verdicts.
• Source URLs must pass a domain allowlist.
• Incident and breach headlines are blocked from becoming tool recommendations.
• ADOPT requires enough readiness evidence or gets demoted.
• Adoption Firewall fails closed on unknown MCP/tool capability surfaces.
• guard never modifies the repo; it only reads local evidence and policy.
• Lab subprocesses receive a stripped environment, wall-clock timeout, size caps, and generated-script secret scanning.
• The import-evidence scanner is deterministic, local, and offline. It parses ASTs via tree-sitter, never sends source content to an LLM, and never reaches the network.

See SECURITY.md for the threat model.

---

💸 Cost

The offline demo is free. A normal live weekly scan is designed to stay cheap:

Component	Typical cost
Sonnet score pass	~$0.15
Sonnet verdict pass	~$0.04
Optional Opus judge	~$0.12
Weekly scan	~$0.30

Set JUDGE_ENABLED=false to skip the Opus judge when you want the cheapest possible run.

---

🗺 Roadmap

• v0.1 — CLI scaffold, local demo, SQLite store, public docs.
• v0.2 — Living Scout Packs, dependency intelligence, Adoption Firewall (evaluate/trial/guard/policy), Incident Change Scout.
• v0.3 — Mission Control terminal setup, provider detection, Scout Pack multi-select, plain/JSON outputs.
• v0.4.0 — Monorepo profile walker + tree-sitter import-evidence scanner (Python and JS/TS), repo-relative manifest_path, --no-imports fast path, .understand-anything/ detection.
• v0.4.1 — Mission Control v2 redesign: branded splash, designer palette, focus borders, modal quit/help/repo-path, RichLog result, sticky status banner, README v2.
• v1.0.0 — Mission Control complete: nine tabs (Scout / Trials / Receipts / Guard / Reports / Packs / Deps / Incident / Settings), scout-first landing with a verdict DataTable and per-verdict actions, every CLI capability has a TUI surface, --tab / --no-scout flags, dismiss persistence.
• v1.1.0 — Global setup wizard (frontier-scout setup), automation mode with cron scheduling, notifications, diff view, Go/Rust/Ruby tree-sitter coverage, frontier-scout doctor, clear-history / notifications / cron run CLI siblings.
• v1.2 — Streaming subprocess output in Trials, multi-repo workspace, PyPI auto-publish on tag.
• v1.3 — launchd / Windows Task Scheduler integrations, live discovery feeds, scout card view.

See ROADMAP.md for the longer view.

---

🤝 Contributing

The fastest useful PRs improve the CLI/report path, validator coverage, source quality, or lab isolation.

• Read CONTRIBUTING.md.
• Browse good first issues.
• Respect the Code of Conduct.

Development loop:

make setup
make demo
make test
make eval
make audit
python -m compileall scripts outputs tests frontier_scout
PYTEST_DISABLE_PLUGIN_AUTOLOAD=1 python -m pytest -q

CI runs compile checks, non-live tests, and a tracked-file secret scan.

Releasing a tagged version

1. Bump project.version in pyproject.toml and frontier_scout/__init__.py.
2. Append a matching ## X.Y.Z - YYYY-MM-DD section to CHANGELOG.md.
3. Merge to main.
4. Push annotated tag vX.Y.Z.

Tag pushes trigger .github/workflows/release.yml, which builds distributions, creates a GitHub Release from the matching changelog section, and (via manual workflow_dispatch) publishes to PyPI via trusted publishing.

---

📄 License

Distributed under the MIT License.

---

🙏 Acknowledgments

• Textual — the framework that makes the Mission Control TUI possible.
• tree-sitter-language-pack — precompiled grammars for the v0.4 import-evidence scanner.
• Pydantic — typed models throughout.
• othneildrew/Best-README-Template — the structure this README borrows from.
• Lum1104/Understand-Anything — the tree-sitter half of its design pushed us to ship deterministic import evidence instead of substring heuristics.

↑ back to top