Precedent Regret — Development skill for Claude Code
Mark a choice you repeated across projects as a mistake, inverting its precedent.
How to install Precedent Regret
Installs to ~/.claude/skills/asm0dey-precedent-precedent-regret/SKILL.md
mkdir -p ~/.claude/skills/asm0dey-precedent-precedent-regret && curl -fsSL https://raw.githubusercontent.com/asm0dey/precedent/HEAD/commands/precedent-regret.md -o ~/.claude/skills/asm0dey-precedent-precedent-regret/SKILL.md Restart Claude Code, or start a new session, for it to be picked up.
What Precedent Regret does
description: Mark a choice you repeated across projects as a mistake, inverting its precedent argument-hint: " "
uv run ~/.claude/skills/precedent/scripts/precedent.py regret \
--topic "$1" --chose "$2" --because "..." --instead "..."
Read `~/.claude/skills/precedent/SKILL.md` first if it is not already loaded.
`--because` is the whole point: without the reason, a future session sees only that the option is disfavoured and cannot tell whethe
Alternatives in Development
- Epic Review — Mark epic review requested, approved, or changes requested 243.5k ★
- Thumbgate Guard — Turn the last agent mistake into a hard prevention rule the agent cannot bypass 26 ★
- SVG Create — Author or clean up an SVG asset — icon, illustration, mascot pose, logo mark — so it scales from viewBox, reco 21 ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
Precedent Check
Show what was decided before on a topic, and flag conflicts with a proposed choice
Detect Pattern Mining
Detect repeated undocumented decisions across session summaries.
Analyze Patterns
Analyze ~/.claude/tool_logs/operations.jsonl operation logs to identify repeated patterns across sessions and
Failure Memory
🧠 Automatic failure pattern recording & recall for OpenClaw agents. Never repeat the same mistake.
Lesson Learned
Write a post-mortem of a mistake into lessons/
Precedent Analyze
Read this project and record the decisions already visible in it
Related Agents
Typesafe Calibrator
Turns invented thresholds into measured ones. Builds labelled sets, plots confidence against accuracy, measure
Mistake Learner
Analyzes mistakes and updates CLAUDE.md with abstracted learnings. MUST BE USED when errors are caught or mist
Design Verifier Xhigh
design-verifier at effort xhigh — for plans or designs where several decisions interlock or a mistake is expen