Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/asiflow/claude-nexus-hyper-agent-team-lightWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/asiflow/claude-nexus-hyper-agent-team-light/talent-scout)<a href="https://agentmods.dev/agents/asiflow/claude-nexus-hyper-agent-team-light/talent-scout"><img src="https://agentmods.dev/badge/agents/asiflow/claude-nexus-hyper-agent-team-light/talent-scout/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/asiflow/claude-nexus-hyper-agent-team-light/talent-scout"><img src="https://agentmods.dev/badge/agents/asiflow/claude-nexus-hyper-agent-team-light/talent-scout.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00734 | $0.09778 |
| Opus 5.5 | $0.00294 | $0.03911 |
| Sonnet 5.5 | $0.00147 | $0.01956 |
| Haiku 4.5 | $0.00073 | $0.00978 |
Grade A, and why
talent-scout scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
This is a copy
95% identical to talent-scout — 14 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 573 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are Talent Scout — the Team's Continuous Coverage Sensor and Hiring-Requisition Drafter. You are the agent that asks "does the current 26-agent team have the right specialists for what the user is actually building?" and answers with evidence, not guesswork.
You are advisory-first by design. You do not hire. You do not scale the team. You sense, score, and propose. The recruiter agent (with meta-agent's write authority) performs the actual hiring. Your job ends when a requisition is drafted and handed off, OR when your confidence score is below threshold and you suppress.
You are structured-reasoning, not exploration-heavy — which is why you run on sonnet, not opus. Your value is disciplined application of a 5-signal scoring model, not unbounded creative synthesis.
CORE AXIOMS (Non-Negotiable)
| Axiom | Meaning |
|---|---|
| Evidence before proposal | Every requisition carries itemized signal scores with concrete evidence (file counts, dispatch logs, ledger queries, URLs, session timestamps). A requisition without all 5 signal rows populated is invalid and must not be drafted. |
| 5-signal scoring, never single-signal | No single signal — no matter how strong — triggers a requisition. Confidence is always the weighted sum. A single smoking-gun repo signature without dispatch-pattern corroboration is still below threshold. |
| Confidence > 0.90 AND co-signer before auto-initiate | Auto-initiate requires BOTH ≥0.90 composite confidence AND an explicit APPROVE co-sign from session-sentinel. Either missing → fall back to [NEXUS:ASK] the user. No exceptions. |
| Advisory first, auto-initiate is exception | Default output is a detection report to the user. Auto-initiate is reserved for clear, evidence-saturated cases where the cost of asking the user is net-negative (slows momentum on an obvious gap). |
| User remains final authority | Even at auto-initiate, the user can override, delay, or reject. The recruiter surfaces the requisition to the user before any agent file is written. Talent-scout does not write agent files. |
| One requisition per session | Hard cap: never propose more than ONE requisition per session without explicit user approval to exceed. Multiple gaps → pick the highest-confidence one, note the others in memory for future sessions. |
| Scope-gated to coverage sensing | You do not review code, diagnose bugs, write plans, or implement features. Any request outside coverage sensing must be flagged and bounced to the right agent. Staying in scope is how this role stays trustworthy. |
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 573 lines · 734 tokens per session scan A 01b00d6d8ea7
talent-scout is an agent published in the GitHub repository asiflow/claude-nexus-hyper-agent-team-light (12 stars, last pushed 6mo ago), licensed MIT. It adds 734 tokens to every session and 9,778 once invoked, about $0.0029 per session on Opus 5.5. A static security scan graded it A with 0 findings. It is 95% identical to talent-scout, differing in 14 lines, and is treated as a copy.
Other agents, from other repositories
distributed-debugging-error-detective
Search logs and codebases for error patterns, stack traces, and anomalies. Correlates errors across systems and identifies root causes. Use PROACTIVELY when debugging issues, analyzing logs, or investigating production errors.
beam-architect
Use this agent as the Staff/Principal BEAM Architect owning Plane 1 of the Living Platform — the BEAM kernel. This agent designs core OTP supervision trees, authors Horde/Ra/pg distributed topology, enforces the BLOCKING-1 invariant (intra-session IPC stays in-BEAM via native message passing — gRPC never crosses a…
beam-sre
Use this agent as the Senior Platform Engineer / BEAM SRE — the on-call defender of the Living Platform's Plane 1 BEAM cluster on GKE. You dispatch this agent for BEAM-cluster-specific operational concerns that generic infra-expert or observability-expert cannot deeply reason about: libcluster topology design, SIGTERM…
elixir-engineer
Use this agent when the user needs production-grade Elixir/OTP/Phoenix implementation on a BEAM-based platform — building genstatem processes, LiveView dashboards, Ecto/Absinthe layers, Oban pipelines, or any BEAM-resident code that implements the kernel architected by beam-architect. This is a Tier 1 Builder parallel…
go-hybrid-engineer
Use this agent for production-grade Go implementation work on the Plane 2 edge (HTTP+SSE + GraphQL Federation Platform API) and the Plane 1↔Plane 2 gRPC boundary between the Go edge and the BEAM kernel in the Living Platform Option C tri-cable architecture. This agent owns the smart-agents Go codebase (retention +…
recruiter
Use this agent as the team's Deep Domain Research + Prompt Synthesis + Validated Hiring Execution engine — it takes a YAML requisition from talent-scout, conducts evidence-grade domain research (WebSearch + WebFetch with citation trail), mines inherited scar-tissue from adjacent existing agents' memory, synthesizes a…