research-target-resolver

Research Target Resolver

Use this skill before company or sector research. Its job is to turn the user's target phrase into a stable ResearchTarget that downstream skills can reuse.

Keep this skill narrow: resolve identity, scope, ambiguity, and initial peer context. Do not write the investment thesis here.

Workflow

1. Parse the user's target phrase, requested market, date range, and intent.
2. Resolve canonical identifiers: company name, ticker, exchange, country, sector, industry, common aliases, and investor-relations site.
3. Detect ambiguity. Ask a concise clarification only if multiple plausible targets would materially change the research.
4. Build a first-pass peer set and sector tags.
5. Create or update the target directory in the research archive.
6. Emit a ResearchTarget block for downstream skills.

Archive Location

Persist reusable target context under:

research/targets/<ticker-or-slug>/profile.md

Use lowercase slugs for non-ticker themes. For public equities, prefer the main listed ticker in lowercase, such as research/targets/intc/profile.md.

Read references/research-target-schema.md before writing a profile or target block.

Output

Return:

• canonical target identity
• ambiguity notes, if any
• peer set and sector tags
• archive path used or created
• source list and freshness notes
• next recommended skills

Quality Gate

Do not hand off to downstream skills until:

• ticker and company identity are explicit for public-company research
• exchange and country are known when available
• aliases include the user's original phrasing
• stale or uncertain facts are labeled
• profile updates preserve useful prior context instead of replacing it blindly