godmode

Jailbreak LLMs: Parseltongue, GODMODE, ULTRAPLINIAN.

Quality

64%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Critical

Do not install without reviewing

Fix and improve this skill with Tessl

tessl review fix ./optional-skills/security/godmode/SKILL.md

Quality

Content

75%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is highly actionable with executable code, well-sequenced workflows including validation and undo, and properly signaled one-level references to real bundle files. Main weakness is length: several inline sections (trigger words, pitfalls, test results) could move to reference files to improve conciseness and progressive disclosure.

Suggestions

Move the 'Trigger Words (Reference)' list and 'Common Pitfalls' into a reference file and link to it, keeping SKILL.md a lean overview.

Relocate the 'Tested Results (March 2026)' section into references or a CHANGELOG, since time-sensitive results age and pad the main body.

Formalize the validation checkpoint in Step 6 as an explicit checklist (validate response score -> escalate -> retry) to push workflow clarity to 5.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient and operational rather than explanatory of basics, but the inlined trigger-word list, 12-item 'Common Pitfalls', and 'Tested Results (March 2026)' section could be tightened or moved to references.	3 / 5
Actionability	Provides copy-paste-ready, executable code and CLI one-liners (load_godmode loader, auto_jailbreak, parseltongue, godmode_race) with concrete examples covering the common cases across all three attack modes.	5 / 5
Workflow Clarity	Steps 0–6 are clearly sequenced with validation checkpoints (baseline test, refusal scoring, dry_run, undo_jailbreak) and explicit escalation/retry loops, though a couple of checkpoints are implicit rather than checklist-formal.	4 / 5
Progressive Disclosure	Real one-level-deep bundle files are clearly signaled ('See references/jailbreak-templates.md', 'See scripts/parseltongue.py', etc.) and confirmed present, but some inlined material (trigger words, model notes) could be split out for cleaner navigation.	4 / 5
	Total	16 / 20 Passed

Description

53%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is terse and names a clear niche plus three concrete techniques, but it omits any 'when to use' trigger guidance and leans on proper-noun jargon over natural user phrasing. Adding a 'Use when...' clause with common synonyms would lift completeness and trigger-term quality.

Suggestions

Add a 'Use when...' clause spelling out natural triggers, e.g. 'Use when bypassing safety filters on Claude, GPT, Gemini, or Grok, or red-teaming a model's safety training.'

Replace or augment the proper-noun technique names with action verbs users actually say ('obfuscate trigger words', 'race models for the least censored response').

Include common synonyms/file cues such as 'uncensor', 'red-team', and 'safety bypass' to improve trigger-term coverage.

Dimension	Reasoning	Score
Specificity	Names the domain ('Jailbreak LLMs') and three concrete named techniques ('Parseltongue, GODMODE, ULTRAPLINIAN'), but lists technique proper nouns rather than concrete actions like 'bypass' or 'obfuscate'.	3 / 5
Completeness	It states a clear 'what' (jailbreak LLMs via three named modes) but has no 'Use when...' or equivalent trigger clause, capping completeness at 3 per the guidelines.	3 / 5
Trigger Term Quality	'Jailbreak' is a natural keyword users say, but the rest are proper-noun jargon few users would naturally utter, and common synonyms ('bypass safety filters', 'red-team', 'uncensor') are missing.	3 / 5
Distinctiveness Conflict Risk	The jailbreaking niche is clearly distinct with named techniques and minimal overlap; only minor overlap risk with the related OBLITERATUS skill noted in metadata.	4 / 5
	Total	13 / 20 Passed

Validation

81%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning
metadata_field	'metadata' should map string keys to string values	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	13 / 16 Passed

Repository: NousResearch/hermes-agent
Path: optional-skills/security/godmode/SKILL.md
Commit: 48f69e5

Reviewed: 1 day ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.