Skip to main content
Glama

PitchSpark: Pre-Seed & Seed Deal Flow

Search pre-seed, seed and early-stage startup pitches with filters, sorting and paging

pitchspark_search_pitches_v2
Read-only

Search founder-submitted startup pitches on PitchSpark with structured filters, five sort orders and cursor paging, returning lightweight cards with the public score, the top two score dimensions and a pass-rate signal (currently withheld pending review). Use when: building or paging through a filtered deal-flow list (stage, raise range, sector, HQ region, traction signals, score floor, recency), or sorting by score, recency, investor views or updates. Prefer it over pitchspark_search_pitches_v1 when you need paging, sorting or recency; pitchspark_search_pitches_v1 stays available unchanged. Do not use for: reading one pitch in depth (pitchspark_get_pitch_v2), score detail (pitchspark_get_score_v1), decision detail from signed-in users (pitchspark_get_decision_summary_v1), founder contact details, decks or data rooms (never returned), investment advice. How search works: rule-based keyword + synonym ranking over the founders' public text; there is NO semantic or embedding search yet, so concrete words work better than abstract ones. query is optional: filters alone list every match. Paging: limit (default 10; the ceiling is 10 without a key or with a free key and 25 with a paid key) and either cursor (from page.next_cursor) or offset; results stop at offset 200. Unlike pitchspark_search_pitches_v1, pitches seen before are not left out (paging needs a stable list). Returns "pitches" (each card: the pitchspark_search_pitches_v1 card keys, plus pitchspark_score (public only, 70+, else null), top_dimensions (the two strongest of the six score dimensions, 0-100, keyed callers only and only when the score is public, else null), decision_signal ({available, reason, decisions_band, pass_rate, min_k}: likes vs passes by signed-in users from stored counters, for investor API keys only (reason key_required otherwise), withheld unless there are at least 5 decisions and the likes and the passes are each 0 or at least 5; the pass rate is currently hidden pending review: reason hidden_pending_review, treat it as no signal), created_at and updated_at), "page" ({offset, limit, returned, total_matches, next_cursor, max_offset}), "sort", "search_mode" ("keyword_synonym"), "interpretation" and "attribution". Repeated identical queries are served from a short-lived cache. Each pitch this tool returns counts 1 against the daily pitch allowance (the same allowance as pitchspark_search_pitches_v1), or 0 when the same API key (without a key, the same connection) already got that pitch today. Calls are logged, and each pitch returned is counted as a view. Recommended flow: pitchspark_search_pitches_v2 (find) → pitchspark_get_pitch_v2 (read) → pitchspark_get_score_v1 (score and roast) → pitchspark_get_decision_summary_v1 (what signed-in PitchSpark users decided) → pitchspark_request_intro_v1 or pitchspark_watch_pitch_v1. Errors are JSON {"error":{"code","message","category","retryable"}}: INVALID_ARGUMENT, UNAUTHORIZED, RATE_LIMITED, INTERNAL.

Tool call template

This tool declares no required parameters. Add optional parameters from the input schema as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {},
    "name": "pitchspark_search_pitches_v2"
  }
}

Input Schema

NameRequiredDescriptionDefault
sortNorelevance (keyword + synonym score; "top scoring" in the query still orders by score), score (public score, pitches without one last), recency (newest listing first), most_viewed_by_investors (Pitch.views_count, all viewers), recently_updated.relevance
limitNoPage size. A value above the caller's tier ceiling is reduced to it.
queryNoOptional plain-English request, 1 to 200 characters, read by the same rules as pitchspark_search_pitches_v1 (sector, stage, place, raise, score and traction cues; quotes and "not X"). Keyword + synonym matching, not semantic search.
stageNoFunding stage or list. "early" means pre_seed, seed and series_a.
cursorNopage.next_cursor from the previous call with the SAME query and filters. Not with offset.
offsetNoSkip this many results instead of a cursor. Not with cursor.
sectorNoSector key or list of up to 10 (primary or additional sector). Keys: fintech, healthtech, saas, marketplace, ai_ml, consumer, enterprise, climate, edtech, crypto, biotech, hardware, gaming, real_estate, food_beverage, fashion, travel, sports, media, femtech, cybersecurity, dev_tools, ecommerce, logistics, mobility, hr_tech, legal_tech, insurtech, proptech, agtech, energy, robotics, govtech, hospitality, other.
locationNoHQ region key or list: us, canada, latam, uk, europe, mea, apac, global_remote.
min_scoreNoPublic PitchSpark score at least this (values under 70 are read as 70; pitches without a public score are left out).
green_flagsNoTraction signal: at least one of these founder-chosen green flags (e.g. "Profitable", "Signed LOI").
has_tractionNoTraction signal: true = only pitches with a public traction note.
max_raise_usdNoRaise at most this many US dollars (disclosed asks only).
min_raise_usdNoRaise at least this many US dollars (disclosed asks only).
created_within_daysNoRecency: listed on PitchSpark within this many days.
updated_within_daysNoRecency: pitch updated within this many days.

Output Schema

NameRequiredDescriptionDefault
pageNo
sortNo
errorNoPresent only when the call failed (isError true).
pitchesNo
truncatedNo
attributionNo
search_modeNoHow text matching works: keyword + synonym ranking. Semantic search is not available yet.
interpretationNo

Schema Changelog

Changes observed during successful MCP inspections.

  1. Added

TDQS

A4.9/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Beyond the readOnly/openWorld annotations, the description discloses the quota model (1 per returned pitch against the daily allowance, 0 for repeat pitches per key), that calls are logged and counted as views, short-lived caching, and that the pass rate is currently withheld pending review with reason 'hidden_pending_review'. It also admits the ranking is keyword+synonym only with no semantic search, and documents the error envelope and codes.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

It is front-loaded and nearly every sentence carries operational value, but the long enumeration of return fields (pitch card keys, page object, sort, search_mode, interpretation, attribution) restates a return shape that the output schema already documents, making the middle dense with duplication an agent does not strictly need.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 15-parameter, open-world search tool this covers everything an agent needs to call it correctly: quota cost, cache behavior, tier-dependent limits, offset cap, cursor/offset mutual exclusion, the withheld pass-rate signal, and a recommended downstream flow. Nothing material is left implicit.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is already 100%, yet the description still adds meaning: query is optional (filters alone list everything), concrete words beat abstract ones given the non-semantic ranker, limit ceilings differ by key tier (10 free / 25 paid), and results stop at offset 200. It also notes that paging does not exclude previously seen pitches, which materially changes how a caller interprets repeated results.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The opening sentence names a specific verb (search), a specific resource (founder-submitted startup pitches on PitchSpark) and the scope of what comes back (lightweight cards with score, top dimensions, pass-rate signal). It also explicitly positions itself against pitchspark_search_pitches_v1 and enumerates the capabilities that distinguish it (paging, sorting, recency).

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It gives an explicit 'Use when' list of filter/sort scenarios, an explicit 'Prefer it over v1 when...' rule with the note that v1 remains unchanged, and a 'Do not use for' list that routes each excluded need to a named sibling (get_pitch_v2, get_score_v1, get_decision_summary_v1). A recommended multi-tool flow is also spelled out.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources