CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

flow-discover

Multi-AI research using available external providers (Double Diamond Discover phase)

Quality

70%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./.claude/skills/flow-discover/SKILL.md

The canonical home for this skill is flow-discover in nyldn/claude-octopus

SKILL.md
Quality
Evals
Security

Quality

Content

81%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.

A tightly written orchestration overview: executable bootstrap code, an unambiguous four-phase sequence with stop conditions and validation gates, and disciplined delegation of detail to one-level-deep reference files. The only real deductions are minor redundancy in the handoff-preservation instructions and the fact that all referenced materials live outside this bundle against a runtime-resolved root.

DimensionReasoningScore

Conciseness

The body is dense host-specific instruction Claude could not know (root resolution, run-ID handoff, security framing) with zero explanation of known concepts, but the literal-assignment preservation requirement is stated three times ("Preserve these printed literal assignments... Include them at the start of each later shell call", "Include its literal assignment in each new shell invocation") and post-compaction reload guidance appears twice. These repetitions could be trimmed, matching anchor 4 rather than 5.

4 / 5

Actionability

Two complete, executable bash blocks (root resolution with exit-on-failure checks; bootstrap invoking "scripts/state-manager.sh handoff_path" and "python3 .../discovery-handoff.py begin") plus concrete named commands ("orchestrate.sh probe-single", "agent-summary", "build-fleet.sh"). Below 5 only because exact per-phase command syntax is deferred to the external reference files, leaving minor gaps in copy-paste coverage here.

4 / 5

Workflow Clarity

A strictly ordered four-phase sequence (prepare, dispatch, synthesize, complete), each bound to a named resource, with explicit validation checkpoints throughout: "Stop on a missing resource", "Wait for all dispatched agents and collect outputs", "Verify a nonempty synthesis", "Do not report completion when dispatch, evidence, or presentation is missing", plus error-recovery guidance ("If the saved values are unavailable, stop and recover the task context") and frontmatter validation gates. This matches the anchor-5 pattern of clear sequencing with explicit validation and recovery.

5 / 5

Progressive Disclosure

The body is a lean overview delegating phase detail to six clearly signaled, one-level-deep references ("Read skills/flow-discover/references/prepare.md", "Load ... guidance.md when selecting context-specific presentation", "Before fetching any external content, read ... security.md"), which is the right pattern. Held at 4 rather than 5 because the referenced paths resolve against an external installation root ($OCTO_ROOT) and no reference files ship in this bundle, so the structure cannot be verified and the reference listing is interleaved with prose rather than gathered in a dedicated navigation section.

4 / 5

Total

17

/

20

Passed

Description

48%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.

A terse, domain-naming description that avoids fluff but undersells the skill: it states no concrete actions, includes no explicit "use when" trigger guidance, and leans on framework jargon ("Double Diamond Discover phase") rather than natural user phrasing. It is distinguishable from other skills but would rarely be selected by natural user intent alone.

Suggestions

Add an explicit trigger clause, e.g. "Use when the user requests multi-provider research or when beginning the Discover phase of a Double Diamond workflow."

List 2-3 concrete actions, e.g. "dispatches parallel research to installed AI providers, collects and validates outputs, and synthesizes cited findings with gap reporting."

Include natural synonyms users would actually say ("multi-model research", "compare AI answers", "research with multiple providers") alongside the framework terminology.

DimensionReasoningScore

Specificity

The description names the domain twice ("Multi-AI research", "Double Diamond Discover phase") but offers only the single generic action "research" — no concrete operations like dispatching to providers, collecting outputs, or synthesizing findings. It is above score 1 (domain is clearly named, not pure abstraction) but below score 3 because it lists no concrete actions at all.

2 / 5

Completeness

The "what" is present but terse ("Multi-AI research using available external providers"), and the "when" is only weakly implied by the parenthetical "(Double Diamond Discover phase)" — there is no "Use when..." clause or equivalent trigger guidance, capping this at 3 per the judging guidelines. Not 4 because the invocation condition is never explicitly stated.

3 / 5

Trigger Term Quality

"Multi-AI research" is a genuinely relevant keyword a user might say, but "available external providers" and "Double Diamond Discover phase" are framework/technical jargon, and common variations ("multi-model research", "compare AI answers", "parallel research") are absent. Fits anchor 3 (some relevant keywords, missing variations/synonyms) better than anchor 2, which requires purely generic or jargon-only language.

3 / 5

Distinctiveness Conflict Risk

"Multi-AI", "external providers", and the phase marker carve a fairly distinct niche with minimal overlap risk against generic research skills. Not 5 because "research" skills are a crowded category and the description lacks distinct explicit trigger phrases to fully disambiguate.

4 / 5

Total

12

/

20

Passed

Validation

87%

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

Validation — 14 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

referenced_paths_exist

Referenced path issues: 2 missing

Warning

Total

14

/

16

Passed

Repository

nyldn/claude-octopus

Path

.claude/skills/flow-discover/SKILL.md

Reviewed

Table of Contents

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.