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 swan-gtm/gtm-skills --skill time-sensitive-sequencegit clone --depth 1 https://github.com/swan-gtm/gtm-skillsWrote 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/swan-gtm/gtm-skills/time-sensitive-sequence)<a href="https://agentmods.dev/skills/swan-gtm/gtm-skills/time-sensitive-sequence"><img src="https://agentmods.dev/badge/skills/swan-gtm/gtm-skills/time-sensitive-sequence/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/swan-gtm/gtm-skills/time-sensitive-sequence"><img src="https://agentmods.dev/badge/skills/swan-gtm/gtm-skills/time-sensitive-sequence.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 | $0.00119 | $0.01155 |
| Opus 5.5 | $0.00048 | $0.00462 |
| Sonnet 5.5 | $0.00024 | $0.00231 |
| Haiku 4.5 | $0.00012 | $0.00115 |
Grade A, and why
time-sensitive-sequence 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 2d 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 — 44 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Applies when a message becomes wrong or useless after a specific moment. Produces sequences plus a scheduled kill that ends them at that moment.
A sequence has no end date. Steps run on waits, and a LinkedIn message can sit for days waiting for a connection to be accepted. Left alone, a pre-event sequence sends "let's meet next week" during event week, and a dinner invite lands after the dinner. The cure has two halves: plan the cadence backwards from the cutoff, and schedule a kill that catches whatever is still alive at the cutoff.
The play
- Set the cutoff. The cutoff is the earlier of two moments: when the copy stops being true, and the last moment an invite is still useful for this kind of event. Read
references/cutoff-rules.mdfor the rules by copy phrase and event type. State the cutoff as a date, time and timezone. - Check what the platform can do. Confirm the six capabilities and look up the seven numbers in
references/platform-assumptions.md. If a capability is missing, use its fallback there, not a guess. - Check feasibility. Work out when the first message can go out and when the last step can run inside the sending window. Divide contacts by daily limits per inbox and add what already waits in the queue. If the list cannot clear before the cutoff, cut the list or add an inbox. Do not build sequences that mostly die in the queue.
- Choose channels. Use both LinkedIn and email when there is time, because they reach different people. Drop LinkedIn when the cutoff is too close for a connection to be accepted and a message to follow. Drop email for people who only respond on LinkedIn. Decide per event and per list, and say why.
- Plan the cadence backwards. Place the last step at least one sending window before the cutoff, with short waits of one to two days. Write copy that is true right up to the cutoff: real dates and times, never "next week" or "this Thursday" unless the cutoff rules say it is safe. Read
references/copy-examples.mdfor real event copy and which phrases in it go stale. - Make one group per event. Three dinners means three groups, three cutoffs, three kill jobs. Never mix cutoffs in one group, because the kill stops the whole group. Put the cutoff in the group's description.
- Build the sequences in the group, in the event's timezone. Leave them for approval unless the owner explicitly said to send automatically.
- Arm the kill job before you call it done. Read
references/kill-job.mdfor the schedule and the hand-off instructions. If you cannot arm it, say so first and do not approve or send. - Report the group, number of sequences, last planned send, cutoff in the owner's timezone and in UTC, and what the kill will and will not touch.
What ships with it
4 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.
- 2d ago First seen · 44 lines · 119 tokens per session scan A 0603dc2bedb6
time-sensitive-sequence is a skill published in the GitHub repository swan-gtm/gtm-skills (181 stars, last pushed 2d ago), licensed MIT. It adds 119 tokens to every session and 1,155 once invoked, about $0.0005 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 2026-10-15.
Other skills, from other repositories
cold-email-personalizer
A tool that writes personalised cold emails, which are unsolicited messages sent to potential customers or partners, using information about each person.
afrexai-lead-hunter
Enterprise-grade B2B lead generation, enrichment, scoring, and outreach sequencing for AI agents. Find ideal prospects, enrich with verified data, score against your ICP, and generate personalized outreach — all autonomously.
campaign-analytics
Analyze campaign performance from deliverability through revenue: delivery, engagement, replies, meetings, SQOs, pipeline, CAC, and payback. Produces diagnostic scorecard, root-cause analysis, benchmark comparison, and next-action plan. Use when deciding whether to scale, fix, or kill campaigns.
gtm-metrics
Build a GTM metrics system for SaaS: funnel conversion, pipeline velocity, win rates, CAC, payback, NRR, sales efficiency, growth accounting, and executive dashboards. Use when creating metrics dashboards, board reports, revenue reviews, operating cadence, or diagnosing growth constraints.
a-b-testing
Design statistically sound GTM experiments for copy, channels, landing pages, sequencing, pricing, and funnel conversion. Produces hypothesis, sample-size logic, success metrics, test plan, analysis method, and scale/stop/kill recommendation. Use when planning A/B tests, split tests, or experiment roadmaps.
attribution
Build practical B2B attribution models across first touch, lead creation, opportunity creation, multi-touch, account-based influence, and sales-sourced revenue. Produces model comparison, UTM governance, source-of-truth rules, and channel ROI view. Use when marketing and sales disagree on source or ROI.