Skip to content
Skills Library
Management & product · Research

explore

/explore

Trigger when user says: "explore options", "what could we do for X", "research approaches", "/empire-research:explore", "investigate approaches", "spawn research team", "what are the options", "options analysis", "explore solutions", "have the team explore". Open-ended exploration: shallow scan enumerates 3–5 candidate approaches, user picks subset to deep-dive, parallel research per approach, consolidated comparison with recommended direction. Findings stay local — never posted externally.

Repository
empire ↗
Licence
MIT
Last repository change
07.10.2026

Assessment and security

92/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content27 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
5
Scripts
1
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 07.10.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What to check before installing
  • 1 scripts in this package. Review them before installing, as Claude may execute them.

Scripts: workflows/explore-deepdive.js

What it does and when to use it

Trigger when user says: "explore options", "what could we do for X", "research approaches", "/empire-research:explore", "investigate approaches", "spawn research team", "what are the options", "options analysis", "explore solutions", "have the team explore". Open-ended exploration: shallow scan enumerates 3–5 candidate approaches, user picks subset to deep-dive, parallel research per approach, consolidated comparison with recommended direction. Findings stay local — never posted externally.

Triggers from the description

explore optionswhat could we do for Xresearch approachesinvestigate approachesspawn research teamwhat are the optionsoptions analysisexplore solutions
Original author description

Trigger when user says: "explore options", "what could we do for X", "research approaches", "/empire-research:explore", "investigate approaches", "spawn research team", "what are the options", "options analysis", "explore solutions", "have the team explore". Open-ended exploration: shallow scan enumerates 3–5 candidate approaches, user picks subset to deep-dive, parallel research per approach, consolidated comparison with recommended direction. Findings stay local — never posted externally.

How it works

SKILL.md has no level-two headings · 115 lines, 907 words

SKILL.md preview
---
name: explore
description: >
  Trigger when user says: "explore options", "what could we do for X",
  "research approaches", "/empire-research:explore",
  "investigate approaches", "spawn research team", "what are the options",
  "options analysis", "explore solutions", "have the team explore". Open-ended
  exploration: shallow scan enumerates 3–5 candidate approaches, user picks
  subset to deep-dive, parallel research per approach, consolidated comparison
  with recommended direction. Findings stay local — never posted externally.
compatibility: Requires network access (web search and fetch); dispatches research subagents. Runs in Claude Code and OpenAI Codex; a bundled persona fills the roster when no named subagents are installed.
---

<section id="purpose-vs-compare">

- Use `explore` when the solution space is open: user knows the problem, not the options
- Use the `compare` skill instead when user already has a known set of options to evaluate head-to-head
- If user input names specific options (A vs B vs C), suggest the `compare` skill and confirm before proceeding here

</section>

<section id="context-gathering">

- Read conversation for problem statement, scope, constraints, success criteria
- Signals to read:
  - Explicit user description of the problem
  - Recent code or files providing technical context
  - Stated constraints (budget, timeline, stack, team size)
  - Prior approaches already ruled out
  - Definition of "good enough" outcome
- If problem statement unclear → ask one clarifying question at a time
- If structured choices help → use `AskUserQuestion` with concrete options
- MUST state inferred problem statement back to user before any dispatch
- MUST get user confirmation on problem statement
- MUST NOT dispatch any agent until problem is confirmed

</section>

<section id="shallow-scan">

- After problem confirmed, dispatch ONE research agent for broad enumeration
- Agent names vary by environment; never assume a specific named agent exists — the bundled persona guarantees the scan can always run
- Two ways to source the scanning agent; prefer the first, always have the second:
  - Named subagent — if the platform exposes a research subagent, pick the best match for general research synthesis or broad information retrieval (Claude Code: the `Agent` tool's `subagent_type`; other agents: their own spawn mechanism)
  - Bundled persona — brief a general-purpose subagent with `references/personas/research-analyst.md`; use when no named subagent fits or the platform has no subagent registry
- Shallow agent instructions:

  - Enumerate 3–5 candidate approaches only
  - One short paragraph per approach — no deep evaluation
  - Required output format:

    ```
    1. <Approach Name>
       <One-paragraph description — what it is, how it addresses the problem>

    2. <Approach Name>
       ...
    ```

  - Cap response under 300 words

- Present shallow-scan output to user verbatim before proceeding

</section>

<section id="user-gate">

- Gate exists because deep-dive spawns one parallel agent per approach (real cost) — user steers spend toward the approaches worth researching
- After shallow scan, present results and ask user:
  - Which approaches to deep-dive (may pick multiple)
  - Whether to add, remove, or reframe any approach
- MUST wait for explicit user selection before deep dispatch
- MUST NOT infer selection and proceed silently
- If user requests a different approach not in list → add it, confirm updated list

</section>

<section id="dispatch-mode">

- After approaches selected, dispatch the deep research one of two ways
- Preferred — a workflow runner is available; the same script runs on every runner, only the call shape differs:

  - Claude Code: `Workflow({ scriptPath: "${CLAUDE_PLUGIN_ROOT}/workflows/explore-deepdive.js", args })`
  - pi: the `pi-dynamic-workflows` extension registers a `workflow` tool — `workflow({ name: "explore-deepdive", description, scriptPath: <this skill dir>/workflows/explore-deepdive.js, args })`; `scriptPath` resolves against the session cwd, so pass the absolute path to the bundled copy; pass the prepared context through the tool's `args` parameter, never baked into the script text. `workflow` appears in the session's native tool set, NOT via MCP. The `workflow` tool is registered lazily: it exists only after `/workflow` has been used in the session (or an eager-registration extension is installed). If `workflow` is absent from the tool set, do NOT search MCP or improvise — ask the user to run `/workflow` once, or use the fallback
  - Any other host exposing a JS workflow runner with `agent()`/`parallel()`: same script, its own call shape
  - The script fans out one researcher per approach with structured pros/cons/fit
  - `args`: `{ problem, constraints, successCriteria, approaches: [{ name, description }] }`

  - Surface the workflow's `log()` lines as progress
  - Feed the returned `approaches[]` into `consolidated-report`
  - The workflow owns dispatch — do NOT also run the inline fallback

- Fallback — ONLY when no workflow runner exists on this host (e.g. OpenAI Codex): read `references/fallback-dispatch.md` and follow its `agent-selection` then `parallel-deep-dispatch` sections

</section>

<section id="consolidated-report">

- After all deep agents return, produce consolidated report
- If the workflow returns `stats.researched < stats.requested`, MUST name the approaches that failed and ask whether to re-run them before presenting
- Comparison table:
  ```
  | Approach | Pros | Cons | Fit |
  |---|---|---|---|
  ```
- `Conflicts` section — where agents cite contradicting evidence; state each side
- `Recommended approach` — prioritized pick with rationale; cite supporting evidence
- MUST cite sources where agents returned citations
- MUST present report then stop; ask user which direction to pursue
- MUST NOT begin implementation

</section>

<section id="guardrails">

- MUST gather and confirm problem context before any agent dispatch
- MUST clarify ambiguity before shallow scan
- MUST confirm shallow results with user before deep dispatch
- MUST dispatch deep research via the `explore-deepdive` workflow when a workflow runner is available; else dispatch deep agents in parallel (single message, multiple tool uses)
- MUST keep all findings local in chat only
- MUST NOT post to Slack, GitHub, Jira, or any external system unless user explicitly authorizes
- MUST NOT implement chosen approach — recommendation only
- MUST NOT proceed through any gate without explicit user confirmation
- The bundled persona in `references/personas/` always provides a fallback researcher — a missing named subagent is never a reason to stop; if the platform cannot spawn subagents at all → MUST stop and tell user; never inline-impersonate a researcher in the main thread

</section>
Files in this package (5)
  • SKILL.md 7 KB
  • references/fallback-dispatch.md 2 KB
  • references/personas/NOTICE.md 1 KB
  • references/personas/research-analyst.md 6 KB
  • workflows/explore-deepdive.js 3 KB

Versions

RepositoryScoreLicence
This page · marcoskichel/empire92MIT
yonatangross/orchestkit100MIT
yonatangross/orchestkit100MIT
rapiddweller/archkeel100MIT
zimoliao/scholaraio100MIT
artk0de/tearags-mcp100MIT
artk0de/tearags-mcp100MIT
artk0de/tearags-mcp100MIT
artk0de/tearags-mcp100MIT

Database snapshot: 13.11.2026 · package version a8f92fe6033a (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 90ideation
    Brainstorm candidate approaches to an open problem. Use when no options are on the table yet, or when a lone approach has never faced a rival. Not for picking among options already on the table (grilling).
  • 100options
    Use when several plans, architectures, or solution approaches need comparing and ranking, or the user asks for options.
  • 100brainstorm
    Divergent-convergent brainstorming with research, ideation, and trade-off analysis. Use when exploring options, generating ideas, comparing approaches. Triggers include "brainstorm", "explore options", "ideate", "compare approaches".
  • 100structured-brainstorming
    This skill should be used when the user needs to brainstorm, explore a problem space, think through design decisions, compare approaches, evaluate trade-offs, is stuck on an approach, or wants to explore multiple solutions. Trigger phrases include 'how should I approach this',…
  • 100evaluate-options
    Research and evaluate implementation options before presenting a recommendation. Use when multiple approaches exist for a problem and a decision is needed. Launches one background research agent per option in parallel, collects evidence-backed findings, then presents a…
  • 100research
    Researches an open-ended question — options, possible solutions, prior art, trade-offs, or how something works — and produces a durable, evidence-backed, adversarially-validated report that recommends an option without committing the team to any artifact. Use when you want to…

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.