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 supermaciz/sessions-chronicle --skill watching-ai-format-evolutiongit clone --depth 1 https://github.com/supermaciz/sessions-chronicleWrote 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/supermaciz/sessions-chronicle/watching-ai-format-evolution)<a href="https://agentmods.dev/skills/supermaciz/sessions-chronicle/watching-ai-format-evolution"><img src="https://agentmods.dev/badge/skills/supermaciz/sessions-chronicle/watching-ai-format-evolution/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/supermaciz/sessions-chronicle/watching-ai-format-evolution"><img src="https://agentmods.dev/badge/skills/supermaciz/sessions-chronicle/watching-ai-format-evolution.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 | $0.00061 | $0.01657 |
| Opus 5 | $0.00030 | $0.00829 |
| Sonnet 5 | $0.00012 | $0.00331 |
| Haiku 4.5 | $0.00006 | $0.00166 |
Grade A, and why
watching-ai-format-evolution 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 9d 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 — 176 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Watching AI Format Evolution
Overview
Run a manual format-watch pass that prioritizes official evidence, extracts only parser-relevant structural signals, and produces a bounded maintenance report. Core principle: separate confirmed upstream evidence from local inference.
Required Output Contract
Unless the user explicitly asks for a different structure, the response MUST be organized as a separate section for each assistant checked or considered. This applies to general monthly watch passes, monthly summaries, watch updates, exploratory checks, and triage requests.
For each assistant section, use these exact headings in this exact order:
Sources checkedConfirmed deltasLikely deltasUnknownsConfidenceRiskInspect nextRecommended action
Do not rename, merge, reorder, omit, or replace these headings.
If a heading has no content, write an explicit placeholder such as None confirmed, Unknown, or Not checked.
The exact heading contract is mandatory. Treat it like a schema, not a stylistic preference.
Use ## <Assistant name> for the assistant section heading, then the exact eight ### headings below it.
Do not answer this skill as freeform prose.
Format Violations
The following are violations of this skill:
- Combining multiple assistants into one shared report
- Using only some of the required headings
- Replacing headings with close variants such as
Evidence,Findings,Next steps, orAction - Writing prose summaries instead of the required per-assistant template
- Using bold assistant names instead of
## <Assistant name>headings - Omitting headings because the answer feels obvious or evidence is thin
Partial compliance is non-compliance.
When to Use
- User asks to watch, monitor, track, or review AI assistant format evolution.
- User wants a periodic manual check on supported or candidate assistants.
- User wants to know whether docs, fixtures, tests, or parser assumptions may be stale.
- User explicitly does not want GitHub Actions or an external bot/service.
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.
- 9d ago First seen · 176 lines · 61 tokens per session scan A 00d073c99b3d
watching-ai-format-evolution is a skill published in the GitHub repository supermaciz/sessions-chronicle (5 stars, last pushed yesterday), licensed MIT. It adds 61 tokens to every session and 1,657 once invoked, about $0.0003 per session on Opus 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 2026-08-31.
Other skills, from other repositories
notification-report
A skill that collects unread notifications from subscribed workspaces into one report. Users can review the report and send replies back to the workspaces that produced the notifications.
teamwork-parent
A workflow for splitting a large issue into smaller sub-issues and coordinating their separate worktrees. A worktree is an isolated copy of a Git project where one task can be developed independently.
background-command
Guidance for running commands that stay active, such as development servers, file watchers, and watch-mode builds, in a separate visible terminal tab. It is written for the oretachi development interface, where users can view logs and stop the process.
presentation
A template-based tool for creating interactive React presentations, meaning slide decks built as web components.
wireframe
A template for an interactive React wireframe, a rough screen design used to plan an interface. It can show several screens, their transitions, and where data fields connect to the interface.
domain-model-diagram
A template for an interactive React domain model diagram, a visual map of the entities and relationships in a software system. It provides grid and focused views with pan, zoom, labels, and category colors.