oracle

Best practices for using the oracle CLI (prompt + file bundling, engines, sessions, and file attachment patterns).

1.70x

Quality

70%

Does it follow best practices?

Impact

99%

1.70x

Average score across 3 eval scenarios

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./openclaw/skills/oracle/SKILL.md

The canonical home for this skill is oracle in Hung-Reo/hungreo-openclaw

Quality

Content

82%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 content is a lean, highly actionable reference for the oracle CLI: concrete copy-paste commands, explicit defaults and edge-case behavior, a golden-path workflow with a dry-run validation checkpoint, and sensible safety guidance. Its weaknesses are minor — the reattach guidance is stated three times, and the ~100-line single file could offload detail sections to reference files.

Dimension	Reasoning	Score
Conciseness	The body is bullet-dense and almost entirely tool-specific knowledge Claude would not have (default-ignored dirs, hidden flags, ~60k-char attachment threshold), assuming Claude's competence throughout. The main redundancy is that "don't re-run; reattach" is repeated across three sections (Main use case, Golden path, Sessions + slugs), which could be trimmed to one.	4 / 5
Actionability	Commands are copy-paste ready and cover the common cases end to end: preview (`oracle --dry-run summary -p "<task>" --file "src/**"`), token sanity, the main browser run with model flags, manual fallback (`oracle --render --copy`), and a full remote host/client pair. The `--file` section shows executable include/exclude patterns, and even hidden aliases (`--copy` for `--copy-markdown`) are documented.	5 / 5
Workflow Clarity	The Golden path is a clearly sequenced 4-step workflow with a pre-flight validation checkpoint (`--dry-run` + `--files-report` to preview payload and token spend) and explicit error-recovery guidance (timeout → reattach to the stored session, don't re-run). Minor gap: the remaining sections are reference-style rather than an ordered procedure, and the prompt-template/checkpoint guidance lacks an explicit verify-output step beyond the one-line "Treat output as advisory".	4 / 5
Progressive Disclosure	The single file is well-sectioned (Commands, Attaching files, Engines, Sessions, Prompt template, Safety) with no nested or buried references, and no bundle files exist so nothing is orphaned. At ~100 lines it exceeds the under-50-line simple-skill case, and detailed material such as engine/remote-host options and the file-attachment default rules could arguably be split into a reference file, keeping it at "good structure with minor organization gaps".	4 / 5
	Total	17 / 20 Passed

Description

58%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 names a distinct tool and enumerates concrete capability areas, but it reads as a table-of-contents rather than an action-and-trigger statement. It lacks any "Use when..." guidance and natural trigger phrasing, so users (and Claude) cannot tell from the description alone when this skill should be invoked.

Suggestions

Add an explicit trigger clause, e.g. "Use when the user mentions oracle, wants to consult another model (GPT/Claude/Grok) with repo context, or asks for a second opinion on code with real files attached."

Lead with third-person action verbs instead of the generic "Best practices for using" framing, e.g. "Bundles a prompt plus selected files into a one-shot request to another model...".

Include natural synonyms and concrete model/tool names users would actually say (e.g. "oracle", "one-shot", "GPT-5.2 Pro", "second opinion") to improve trigger-term coverage.

Dimension	Reasoning	Score
Specificity	The parenthetical lists several concrete capability areas — "prompt + file bundling, engines, sessions, and file attachment patterns" — but the framing "Best practices for using the oracle CLI" is generic, and the items are topics rather than actions, so it falls between anchors 3 and 4, noticeably above the midpoint.	4 / 5
Completeness	It clearly answers "what" (best practices for prompt + file bundling, engines, sessions, attachments) but there is no "Use when..." clause or equivalent trigger guidance, which caps completeness at 3 per the judging guidelines.	3 / 5
Trigger Term Quality	"oracle CLI" and "file attachment" are relevant keywords, but terms like "engines", "sessions", and "prompt + file bundling" are technical jargon rather than natural phrases a user would say; it misses common variations such as "ask another model", "second opinion", or the target model names.	3 / 5
Distinctiveness Conflict Risk	"oracle CLI" names a distinct niche tool, so conflict risk is low, but the abstract "best practices for using" framing without trigger phrases leaves minor overlap risk with generic CLI-usage skills — mostly distinct rather than a clear fully-specified niche.	4 / 5
	Total	14 / 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: trpc-group/trpc-agent-go
Path: openclaw/skills/oracle/SKILL.md
Commit: 9bb1c42

Reviewed: 5 days 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.