Borrowing it
Nothing to install: this file belongs to jahala/umbel. 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/jahala/umbel/master/.claude/skills/tend-change/SKILL.mdgit clone --depth 1 https://github.com/jahala/umbelWrote 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/jahala/umbel/tend-change)<a href="https://agentmods.dev/skills/jahala/umbel/tend-change"><img src="https://agentmods.dev/badge/skills/jahala/umbel/tend-change/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/jahala/umbel/tend-change"><img src="https://agentmods.dev/badge/skills/jahala/umbel/tend-change.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 | $0.00051 | $0.05150 |
| Opus 5 | $0.00026 | $0.02575 |
| Sonnet 5 | $0.00010 | $0.01030 |
| Haiku 4.5 | $0.00005 | $0.00515 |
Grade A, and why
tend-change 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 10d 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 — 376 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Change Request: Update + Cascade
Handle changes to what a feature does, requires, or connects to. Updates the affected feature and cascades impact assessments - to step traces, to dependent features, and to audit results that may now be stale.
What this skill assumes
- The change has been identified — the user has either described it or this skill is invoked from a context where the change is implicit (e.g., a brainstorm or audit run surfacing the need).
- The project state on disk is consistent (no half-applied edits from a previous interrupted change).
- For spine-level changes (renaming a persona, reordering
jobs[], editingopportunity.personas), the user provides an explicit reindex/rename map when this skill asks for one. The skill refuses silent corruption. - Audit cascade is the user's responsibility post-change. This skill identifies what needs re-verification; the user runs
/tend auditto close the loop.
Why this matters
When checks change, every step that traces to them and every audit verdict against them becomes potentially invalid. The MCP write's auto-side-effects handle the easy cases (audit auto-invalidates on spec-of-truth field changes), but human-judgment cascades to dependent features are this skill's job. Without explicit handling, you end up with steps tracing to deleted checks, dependents that silently assume edges that no longer exist, and audits whose verdicts contradict the checks they reference.
Rules
- Always show a before/after diff for every changed field before writing. Confirm with the user before applying.
- Ask confirmation before making any changes to dependent features - the user decides cascade scope.
- Ask the user for clarification when a change is ambiguous - do not guess intent.
- Do not auto-rewrite checks on dependent features - flag them as needing review, never silently change them.
- If a change removes a dependency edge, check whether any dependent feature assumed that edge was satisfied - flag it.
- After all updates, run
tend_validateand surface any schema errors (especially orphan checks after a step trace is removed). - If the change would create a dependency cycle, refuse and explain why.
- Audit auto-invalidation is handled by the MCP write. Any material change to a check (
descriptionmaterial change,validates,validates_job,method,integration_level,verification_recipe), any change to a step'sdescription/traces_to, or any change tosmoke.cmdcauses the MCP write to clearauditand dropstatusfromverifiedtoin-progressin the same call. The skill does not call invalidation explicitly - the side-effect is what makes the cascade durable. - The skill does the human-judgment cascade. MCP handles "this feature's audit is stale". The skill handles "feature X is in feature Y's dependencies, and X just changed its surface area - does Y need to revisit its checks?"
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.
- 10d ago First seen · 376 lines · 51 tokens per session scan A 899fd4782aa0
tend-change is a skill published in the GitHub repository jahala/umbel (2 stars, last pushed 4d ago), licensed MIT. It adds 51 tokens to every session and 5,150 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
factory-handoff
Hand one exact agent-ready ticket to the central Factory runtime over the configured SSH forced-command boundary. Use when an operator asks to hand off an existing ticket or a prose request instead of implementing it in the current session.
ticket-spec
Turn a raw Linear Triage ticket into a fully specified ai:agent-ready one by exploring the codebase. Use when specifying, refining, or promoting tickets, or when asked why a ticket isn't dispatchable.
swarm-plan
Full execution protocol for MODE: PLAN -- plan creation, external plan ingestion, QA gate persistence, task granularity, and traceability checks.
loop
Full execution protocol for MODE: LOOP — the compound-engineering loop: brainstorm → plan → build → review → improve, iterating under defense-in-depth stop conditions with generator/critic separation, durable resumable state, and mandatory compounding learning capture. Loaded on demand by the architect when the loop…
issue-ingest
Full execution protocol for MODE: ISSUEINGEST -- GitHub issue intake, localization, spec generation, and transition to the full fix workflow.
phase-wrap
Claude Code adapter for MODE: PHASE-WRAP. Delegates to the canonical opencode-swarm phase boundary protocol.