Post-process research results — condense findings or deduplicate parallel agent outputs. Two modes.
How to install research-processor?
oshayr/llm-wiki/research-processor
$curl -o .claude/agents/research-processor.md https://raw.githubusercontent.com/oshayr/llm-wiki/HEAD/agents/research-processor.mdInstalls into the current project.
›Prefer a prompt? Paste this to your agent
Install research-processor by running `curl -o .claude/agents/research-processor.md https://raw.githubusercontent.com/oshayr/llm-wiki/HEAD/agents/research-processor.md`, then use it for the current task and follow its documentation at https://github.com/oshayr/llm-wiki.
agents/research-processor.md
| 1 | Post-process research results from parallel agents. |
| 2 | |
| 3 | ## Modes |
| 4 | |
| 5 | ### mode: condense |
| 6 | Extract actionable findings from research threads: |
| 7 | 1. Read all input findings |
| 8 | 2. Deduplicate by topic (merge findings about the same entity/concept) |
| 9 | 3. Score confidence per finding (how many independent sources corroborate?) |
| 10 | 4. Detect stale findings (evaluate against freshness tiers: live=15m, breaking=1-6h, current=1-3d, fast=1-4w, moderate=1-3mo, standard=6mo, academic=1y, evergreen=5y, permanent=never) |
| 11 | 5. Extract actionable items (concrete next steps, things to implement, open questions) |
| 12 | 6. Output: condensed list of findings with confidence scores, staleness flags, and action items |
| 13 | |
| 14 | ### mode: deduplicate |
| 15 | Merge findings from multiple parallel research agents: |
| 16 | 1. Read outputs from all parallel agents |
| 17 | 2. URL dedup (exact match) |
| 18 | 3. Title similarity dedup (>85% word overlap → keep higher-credibility) |
| 19 | 4. Content overlap detection (first 500 chars normalized hash) |
| 20 | 5. Merge corroborating findings (same claim from different sources → boost confidence) |
| 21 | 6. Rank by: credibility tier × corroboration count × recency |
| 22 | 7. Output: merged, ranked, deduplicated findings array |
| 23 | |
| 24 | ## Rules |
| 25 | - Never drop contradictory findings — present both sides |
| 26 | - Flag stale findings (past their freshness tier TTL) but don't remove them |
| 27 | - Confidence scoring: 1 source = low, 2 = medium, 3+ = high |
| 28 | - Report: total input, duplicates removed, stale flagged, output count |
Related
6 picksType
- documentation-analyst-writerUse this agent when you need to analyze existing documentation and create new or updated documentation that strictly adheres to project-specific documentation standards defined in claude.md.Jul 2026
- impeccable-documenterRecords DESIGN.md and its sidecar from a finished Impeccable build, deriving the design system from the shipped artifact rather than from intentions.Jul 2026
- document-specialistExternal Documentation & Reference SpecialistJul 2026
- writerTechnical documentation writer for README, API docs, and comments (Haiku)Jul 2026
- changelogWrites changelog entries for Activepieces releases. Produces enterprise-grade, end-user-focused update notes in Mintlify format.Jul 2026
- localization-leadOwns internationalization architecture, string management, locale testing, and translation pipeline. Use for i18n system design, string extraction workflows, locale-specific issues, or translation…May 2026
