Install
$ agentstack add skill-tmusser-ai-engineering-skills-ceremony-budget ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
About
Ceremony Budget
Purpose
Route a task to the smallest safe workflow before the agent starts creating artifacts or editing files.
When to use
Use at the start of a task when it is not obvious whether the work needs no artifact, a short route, or a fuller workflow.
Inputs
- User request
- Known constraints
- Expected blast radius
- Verification burden
- Resume or handoff risk
- Decision or irreversibility risk
Workflow
- Check whether the task is reversible, bounded, and already unambiguous.
- Check whether failure would mostly waste time or would affect users, data,
decisions, compatibility seams, or future sessions.
- Check whether the task is likely to drift without a scope boundary or explicit
verification target.
- Choose the lowest level that still protects against the real failure mode.
- Select the skill route for that level.
- Name the skills or artifacts intentionally skipped.
- Reserve the minimum proof needed before stopping.
- Stop routing once the decision block is good enough to start the work.
Outputs
Produce a short routing block. Do not create a durable file by default.
CEREMONY BUDGET
Level: 0 | 1 | 2 | 3
Why: ...
Use: ...
Skip: ...
Proof reserve: ...
Stop rule: ...
Level guide:
- Level 0: tiny reversible patch; usually no durable artifact
- Level 1: one bounded slice; inline boundary plus proof
- Level 2: small multi-step or ambiguity-prone slice; compact spec and
verification
- Level 3: decision-impacting, autonomous, user-facing, risky, or multi-slice
work; full guarded route
Escalate when:
- ambiguity blocks implementation
- scope is starting to drift
- verification is weak or hand-wavy
- context is likely to be lost before completion
- another session must resume safely
- changes are hard to undo or affect decisions, users, or public seams
De-escalate when:
- the task is a tiny reversible edit
- the route is creating artifacts that are not buying clarity or safety
- one shorter proof path would cover the same risk
- the next skill would only restate what is already explicit
Stop conditions
- The lowest safe level is chosen.
- The selected route and proof reserve are explicit.
- Do not continue routing once the next step is clear.
Anti-patterns
- Treating more artifacts as automatically better process.
- Creating
SPEC.md,PLAN.md, orHANDOFF.mdfor a trivial reversible fix. - Creating
CEREMONY_BUDGET.mdby default. - Escalating because the repo has a skill available, not because the task needs it.
- Using Level 0 or 1 when the task is decision-impacting or hard to resume safely.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: tmusser
- Source: tmusser/ai-engineering-skills
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.