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.
npx skills add GTMify/aigtm --skill agent-orchestrationgit clone --depth 1 https://github.com/GTMify/aigtmWrote 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/skills/gtmify/aigtm/agent-orchestration)<a href="https://agentmods.dev/skills/gtmify/aigtm/agent-orchestration"><img src="https://agentmods.dev/badge/skills/gtmify/aigtm/agent-orchestration/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/skills/gtmify/aigtm/agent-orchestration"><img src="https://agentmods.dev/badge/skills/gtmify/aigtm/agent-orchestration.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 | $0.00100 | $0.01770 |
| Opus 5.5 | $0.00040 | $0.00708 |
| Sonnet 5.5 | $0.00020 | $0.00354 |
| Haiku 4.5 | $0.00010 | $0.00177 |
Grade A, and why
agent-orchestration 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.
How it starts
The opening of the file, as written. The whole thing — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent Orchestration Agent
Your Role
You are a GTM systems architect. Your job is to take skills the operator already has and wire them into a single motion that mostly runs itself, while keeping the human firmly in control of anything that touches a customer or cannot be undone. You do not invent new capabilities. You compose the skills already installed in this library into a chain where each step's output becomes the next step's input, and you place a deliberate approval gate before any irreversible action.
This is the sequel to the Power Prompting move of saving a prompt as a skill with a trigger phrase. There the unit of work was one skill. Here the unit of work is a chain of skills that a system can run start to finish, supervised by a person at exactly one well-chosen point.
The Five-Layer Autonomous GTM Stack
Every orchestration you design maps to five layers. Name them out loud when you plan a motion so the operator can see the shape of the system.
- Trigger. What kicks the motion off. A new batch of target accounts, an inbound form fill, a CRM stage change, a Monday morning schedule. The trigger defines the input the first skill receives.
- Agent / decision. The reasoning step that reads the input and decides what to do. This is where a research or scoring skill turns raw input into a qualified, prioritized set of actions. The decision layer is allowed to say "none of these are worth pursuing," and that is a valid, good outcome.
- Execution / tools. The muscle. The step that produces the actual work product: drafted emails, a sequence, a summary, a document. Execution drafts; it does not send.
- Human-in-the-loop gate. A required checkpoint before anything customer-facing or irreversible. The operator reviews, edits, approves, or kills the batch. Nothing crosses this line without an explicit approval. This layer is not optional and it is not a formality.
- Feedback. What the motion learns from. Outcomes flow back into the operator's Context (Layer 1 of the Power Prompt Stack), the ICP, and the messaging, so the next run of the motion is sharper than the last.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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 · 76 lines · 100 tokens per session scan A d24f202e4e07
agent-orchestration is a skill published in the GitHub repository GTMify/aigtm (25 stars, last pushed 23d ago), licensed MIT. It adds 100 tokens to every session and 1,770 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
activity-capture-diagnostic
Diagnose sales activity capture quality across a team and grade it A through F. Computes per-rep capture rate, under-loggers below team median, phantom-progression opportunities (stage advanced with zero logged activity), activity type mix, and the correlation between capture rate and win rate. Use when the user asks…
lead-routing-rule-analyzer
Diagnose lead routing health and grade it A through F. Computes speed-to-lead (creation-to-assignment and assignment-to-first-touch) with percentiles, SLA violations, orphaned leads, distribution imbalance across reps, and routing leakage (assigned rep territory does not match lead territory). Use when the user asks…
salesforce-revops-audit
Run a comprehensive Salesforce RevOps health audit across five dimensions (pipeline health, data quality, forecast hygiene, deal integrity, process integrity) and produce a graded report (A through F) with a prioritized remediation queue. Use as the entry point for any new Salesforce org assessment, before a new…
deal-investigator
Produce a structured deal review for a single opportunity. Pulls activity, stage progression, comparable deals, and risk indicators. Use when a manager needs to pressure-test a specific deal, prepare for a deal review, decide whether to escalate, or determine if a deal is real before the forecast call.
forecast-call-prep
Prepare the executive summary for a weekly forecast call by reconciling rep commits against historical close-rate models and surfacing material variance. Use when the user asks for forecast call prep, weekly forecast review, forecast variance analysis, or a Monday morning forecast memo.
pipeline-hygiene-audit
Audit a sales team's open pipeline for hygiene issues and produce a per-rep punchlist of fixes ranked by severity. Use when the user asks for pipeline hygiene, data quality check, stale opp report, pipeline cleanup, or CRM cleanup before a forecast call or QBR.