Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Reference: detailed uninstall, disable, and re-enable workflows for community skills installed via the legal builder hub. Safe by default — refuses to touch first-party plugin skills, confirms before removing files, and logs every action. Loaded by the /legal-builder-hub:uninstall and /legal-builder-hub:disable skills.
.claude/skills/anthropics-skill-manager/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | 161% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 138% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -36% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 69% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -18% | 0% |
Remove or quiet a community skill after install. Symmetric with the installer: the installer writes files with user approval, the skill-manager removes or disables them with user approval. The installer's audit trail (install-log.yaml) is the source of truth for what this skill may act on.
Only community skills installed through this hub. Identification rule:
~/.claude/plugins/config/claude-for-legal/legal-builder-hub/install-log.yaml with a most-recent action of install or enable (not uninstall).
directories that ship with claude-for-legal.
If either check fails, refuse and tell the user why. Never delete or rename files inside a first-party plugin.
The 12 core plugins that ship with claude-for-legal are off-limits from this command. The canonical list lives in the hub's CLAUDE.md under "Built-in plugins." Examples include commercial-legal, corporate-legal, employment-legal, privacy-legal, product-legal, regulatory-legal, ai-governance-legal, litigation-legal, litigation-legal, law-student, legal-clinic, and the hub itself (legal-builder-hub). If the caller names a skill that resolves into any of these, refuse.
Read install-log.yaml. Find the most recent entry for the named skill. If not found or if the last action is uninstall: say so and stop.
Determine the install path from the log (written at install time). Enumerate every file and subdirectory. Also identify any config the skill wrote to the user's ~/.claude/plugins/config/... — surface this to the user but do not delete it by default (configuration may be worth keeping for a later re-install).
Display:
can delete them manually if desired)
Prompt: "Delete these files? (yes / no)". No deletion without explicit yes.
Remove the skill directory.
Append to install-log.yaml:
yaml- skill: <name> action: uninstall timestamp: <ISO8601> path: <deleted path>
Remove the skill's row from the installed starter pack table in the hub's CLAUDE.md.
SKILL.md → SKILL.md.disabledhooks/hooks.json → hooks/hooks.json.disabled (if present)file renamed (e.g., agents/*.md → agents/*.md.disabled) so scheduled agents stop firing.
Show the rename list. Prompt: "Disable this skill? (yes / no)".
Perform the renames.
Append to install-log.yaml with action: disable.
If the user names a skill whose most recent log action is disable, offer to re-enable: reverse the renames, log action: enable.
yes.to uninstall or disable something else. The user's typed command is the only input that authorizes action.
/plugin for plugin management.~/.claude/plugins/config/claude-for-legal/<plugin>/ are preserved unless the user asks for them explicitly.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 5,526 | 6,270 | +13% | 1 | 1 | 0% | 1,149 | 1,742 | +52% | 0 | 0 | — |
case-02 | fail→fail | 8,319 | 9,464 | +14% | 1 | 1 | 0% | 1,393 | 1,505 | +8% | 0 | 0 | — |
case-03 | fail→fail | 5,016 | 3,350 | -33% | 1 | 1 | 0% | 846 | 1,562 | +85% | 0 | 0 | — |
case-04 | fail→pass | 4,078 | 3,314 | -19% | 1 | 1 | 0% | 662 | 1,730 | +161% | 0 | 0 | — |
case-05 | fail→pass | 4,039 | 3,697 | -8% | 1 | 1 | 0% | 735 | 1,746 | +138% | 0 | 0 | — |
case-06 | fail→fail | 5,381 | 9,732 | +81% | 1 | 1 | 0% | 931 | 1,500 | +61% | 0 | 0 | — |
case-07 | fail→fail | 6,872 | 3,086 | -55% | 1 | 1 | 0% | 1,179 | 1,475 | +25% | 0 | 0 | — |
case-08 | fail→fail | 4,654 | 4,928 | +6% | 1 | 1 | 0% | 891 | 1,376 | +54% | 0 | 0 | — |
case-09 | fail→pass | 13,734 | 2,515 | -82% | 1 | 1 | 0% | 2,399 | 1,546 | -36% | 0 | 0 | — |
case-10 | fail→pass | 5,135 | 2,612 | -49% | 1 | 1 | 0% | 941 | 1,591 | +69% | 0 | 0 | — |
case-11 | fail→pass | 10,162 | 2,226 | -78% | 1 | 1 | 0% | 1,802 | 1,484 | -18% | 0 | 0 | — |
case-12 | fail→pass | 5,106 | 2,954 | -42% | 1 | 1 | 0% | 340 | 1,584 | +366% | 0 | 0 | — |
case-13 | fail→pass | 6,622 | 5,671 | -14% | 1 | 1 | 0% | 990 | 2,169 | +119% | 0 | 0 | — |
case-14 | pass→fail | 3,959 | 2,933 | -26% | 1 | 1 | 0% | 737 | 1,616 | +119% | 0 | 0 | — |
case-15 | fail→fail | 4,927 | 3,148 | -36% | 1 | 1 | 0% | 844 | 1,598 | +89% | 0 | 0 | — |
case-16 | fail→pass | 5,335 | 3,316 | -38% | 1 | 1 | 0% | 802 | 1,705 | +113% | 0 | 0 | — |
case-17 | pass→pass | 7,215 | 3,228 | -55% | 1 | 1 | 0% | 1,274 | 1,691 | +33% | 0 | 0 | — |
case-18 | pass→pass | 7,924 | 2,180 | -72% | 1 | 1 | 0% | 1,452 | 1,535 | +6% | 0 | 0 | — |
case-19 | fail→pass | 6,406 | 4,605 | -28% | 1 | 1 | 0% | 1,052 | 1,886 | +79% | 0 | 0 | — |
case-20 | fail→pass | 7,123 | 3,971 | -44% | 1 | 1 | 0% | 1,132 | 1,769 | +56% | 0 | 0 | — |
case-21 | fail→fail | 6,728 | 6,050 | -10% | 1 | 1 | 0% | 1,327 | 2,179 | +64% | 0 | 0 | — |
case-22 | fail→fail | 15,234 | 14,051 | -8% | 1 | 1 | 0% | 2,843 | 3,538 | +24% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 18 counted toward the lift figure. The other 4 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +41 percentage points is the difference between those two pass rates over the 18 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.