Borrowing it
Nothing to install: this file belongs to seed-hypermedia/seed. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/seed-hypermedia/seed/main/.agents/skills/weekly-demo/SKILL.mdgit clone --depth 1 https://github.com/seed-hypermedia/seedWrote 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/seed-hypermedia/seed/weekly-demo)<a href="https://agentmods.dev/skills/seed-hypermedia/seed/weekly-demo"><img src="https://agentmods.dev/badge/skills/seed-hypermedia/seed/weekly-demo/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/seed-hypermedia/seed/weekly-demo"><img src="https://agentmods.dev/badge/skills/seed-hypermedia/seed/weekly-demo.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 | $0.00039 | $0.01480 |
| Opus 5.5 | $0.00016 | $0.00592 |
| Sonnet 5.5 | $0.00008 | $0.00296 |
| Haiku 4.5 | $0.00004 | $0.00148 |
Grade A, and why
weekly-demo 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 20d 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 — 99 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Weekly demo
Turn the user's recent repository work into a meeting guide, not a commit dump. Explain what changed, why it matters, and what to show. Include unfinished local work without presenting it as shipped or verified.
Invocation and defaults
Examples:
$weekly-demo— last seven days, plus current local changes.$weekly-demo last week, 5-minute demo— previous Monday through Sunday in the user's timezone.$weekly-demo since 2026-10-01, 10 minutes— explicit start date through now.$weekly-demo refresh— update the previous report using the same dates and author scope.
Accept dates, an author, a branch, and a meeting duration in ordinary language. Do not require flags or a script. Default to a five-minute demo. State the exact date range, timezone, author identities, branch, and review time. Do not silently expand the period when little work is found. Date-only end dates include that full local day.
Use the current repository and worktree by default. If the user requests main while another branch is active, inspect the local main ref without switching branches. Label current-worktree edits separately. Do not include other branches or worktrees unless requested. Ask only when identity or scope cannot be resolved with reasonable confidence.
Collect evidence without changing the repository
Follow repository instructions. In Seed, run repository commands with direnv exec .. Do not fetch, switch branches,
stage, commit, alter files, or start an app to prepare this report. Do not create a report file unless requested. Do not
read ignored files, credentials, or private security notes.
- Read Git status, the current branch, local refs, remotes, and configured user name/email. Check for a shallow clone
or missing refs and disclose incomplete history. Inspect recent author identities and
.mailmap, if present. Include confirmed aliases for the same person; do not match unrelated people by a loose substring. Do not count other people's commits merely because they are on the user's branch. Label explicit co-authored work as shared. - Read all commits in the interval on the selected ref, including full hashes, author and committer dates, subjects,
and bodies. Use committer dates for the reporting interval and disclose that convention. Use
--since-as-filterwith explicit timezone-aware start and end timestamps so unusual date ordering does not hide commits. Inspect other authors' changes only as needed to establish the final state of the user's work. - Read both staged and unstaged diffs and list non-ignored untracked files. Inspect relevant source and tests, including untracked source files, rather than relying only on diff statistics. Do not print private data or large generated/binary files. Local changes have no reliable author or work date: label them as current local work, outside the commit date filter, with authorship unverified.
- For each candidate demo, inspect commit details and current code as needed to establish the user-visible result. Combine related fixes, polish, tests, and follow-up commits into one feature story. Check for reverts, disabled features, superseded changes, and work that is no longer present. Put these in a short caution section, not the demo.
- Compare the selected ref with its configured upstream when available. Distinguish commits present in that local upstream ref from local-only commits. Label remote information as a local snapshot; do not claim a live remote check. No upstream means unknown sync status, not that every commit is unpushed.
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.
- 20d ago First seen · 99 lines · 39 tokens per session scan A 7f8b3a4428e7
weekly-demo is a skill published in the GitHub repository seed-hypermedia/seed (59 stars, last pushed today), licensed Apache-2.0. It adds 39 tokens to every session and 1,480 once invoked, about $0.0002 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 2026-10-09.
Other skills, from other repositories
gsd-quick-batch
Batch several /skill:gsd-quick-shaped tasks together — planned, dispatched, and merged as one run.
gsd-pause-work
Create context handoff when pausing work mid-phase.
mcore-split-pr
Split a PR into multiple PRs to reduce the number of required CODEOWNERS reviewer groups.
engineer-bot-ops
A workflow for carrying an engineering task from scope through implementation and reviewable delivery. It works with a repository's branch, existing changes, rules, checks, and permissions while separating completed, unverified, and next-step work.
gait-session-tracking
GAIT session lifecycle management - branch creation, turn recording, audit logging for every NetClaw operation. Use when starting a new NetClaw session, recording a health check or config change, pinning a pre-change baseline, or viewing the audit trail for a troubleshooting session.
close-task-commit-push-pr
Close the active backlog task (detected from branch name), commit all changes, push to remote, and open a pull request. Use when the user says "close task and ship it", "close task commit push pr", or invokes /close-task-commit-push-pr.