agenthropic

Market Landscape & Selection Bias

The panel’s blind spot

The reports frame the market as “only tiny single-maintainer projects; no Grafana of Claude Code exists.” That framing is broken by popular, MIT-licensed, actively maintained tools the reports never mention:

Missed tool	Stars	What it is	Why it still isn’t the answer
davila7/claude-code-templates	28.4k, MIT	Real npx claude-code-templates --analytics live web dashboard; reads ~/.claude JSONL; zero-install; no Docker	Flat subagent leaderboard, no DAG; no dollar cost; no persistence (in-memory TTL cache); no Telegram; binds 0.0.0.0
jarrodwatts/claude-hud	26.1k, MIT	Literally “see which subagents are running”	In-terminal statusline HUD; single-session; not a historical web cockpit
ccusage	16.8k, MIT	The de-facto token/cost tool for Claude Code	CLI report, not a cockpit — but it undercuts the report’s praise of cast as capturing cost “no one else does”

See projects/claude-code-templates.md for the full deep-dive on the most relevant of these.

Honest reading of the thesis

• As literally written — “no feature-rich AND multi-maintainer AND >1k★ option exists” — the thesis is false. claude-code-templates is all three.
• Narrowed — “no DAG cockpit (per-instance orchestration graph + dollar cost + persistence + alerting) that is also multi-maintainer and popular” — the thesis survives. That is the real, defensible gap.

The market gap that justifies building

No tool in the surveyed set — including the popular ones the panel missed — delivers all of:

1. Global, persistent, per-instance subagent orchestration DAG (not a session-scoped, render-time-derived tree).
2. Dollar-cost attribution + delegation-savings, surfaced live.
3. Telegram alert sink (→ @baev_bot_bot).
4. Cross-machine / fleet aggregation.
5. Persistence you own for historical/time-series analysis.

claude-code-templates is the honest baseline to differentiate against: it already nails self-hosted + zero-install + live token attribution. The OPCⁿ moat is precisely the DAG + dollar-cost + persistence + Telegram it lacks. Building agenthropic is justified — but the pitch is “the persistent DAG cockpit,” not “the first Claude Code dashboard.”

Implication for agenthropic

• Don’t position against a vacuum; position against claude-code-templates’ feature set and beat it on the five gaps above.
• Reuse its proven pattern (chokidar watch over ~/.claude JSONL, zero-install DX) as a UX bar to clear.

This site is open source. Improve this page.