Issuers where a holder has a big stake AND currently sits on the board.

The strongest single output here, and the best place to start a governance or counterparty question. It is the two-limb test from 12 CFR 225.2(e) — the Federal Reserve's control presumptions — and its force comes from the conjunction of two independent filing types: a Schedule 13D on one side, Form 3/4/5 board activity on the other. A screener can give you either list; the pairing is the finding.

Requiring a current board seat also fixes a freshness problem. 13D carries no exit obligation below 5%, so half the stakes on file predate 2020 and are last-known rather than current — but board activity runs to the present, so a recent seat corroborates an old declaration. Liberty Broadband's 26.1% of Charter was declared in 2014; its director was seen in 2026.

min_tier is a presumption tier (10/15/25/50). min_value_usd filters on the 13F size proxy so results are recognizable names rather than nano-caps.

Caveat to state if asked: percent_of_class is percent of the class covered by the filing, NOT voting power — and several of the largest names are dual-class (Berkshire, the Liberty complex, Carvana, Sea), where economic and voting stakes diverge.

Example: influence_map() returns Buffett/Berkshire 37% (seat 2026-05), Liberty Broadband/Charter 26.1% (seat 2026-06), Huffman/Reddit 61.5% (seat 2026-06).

Screen for issuers where several known activist franchises filed 13D close together.

The "what is heating up" screen, and the best starting point for an event-driven question. since is an ISO date (e.g. '2025-01-01'). Returns each issuer with the franchises involved, the span in days, and the filing sequence with accession numbers.

Gated to a curated list of recognised activist franchises (Icahn, GAMCO/Gabelli, Saba, Bulldog, Karpus, Elliott, Starboard, RA Capital, OrbiMed, ...) — precision over recall. Without that gate the output is dominated by micro-cap founders crossing 5% and by filing-group artifacts where one manager files through several affiliated entities.

Example: activist_convergence(since='2023-01-01') surfaces MNRO (GAMCO then Icahn, 96 days apart) and SION (OrbiMed then RA Capital, 5 days apart).

Show who moved first on an issuer, who followed, and who is merely index money.

Returns every dated 13D/13G filing on the issuer in chronological order, each filer classified as activist / insider_or_other / passive_index / custodian, with percent_of_class and the accession number. Also reports the first mover and, for each follower, how many days later they arrived. Abstains when the issuer has no dated ownership filing.

Only 13D/13G filing dates are used — they are the one trustworthy time series in this graph (board and 13F layers are snapshots).

Example: campaign_timeline("MNRO") shows GAMCO at 4.0% on 2025-08-01, then ICAHN CARL C at 14.79% exactly 96 days later.

Trace the transitive >=50% ownership control chain through a public issuer.

direction='up' = who ultimately controls this issuer; 'down' = what this issuer controls. Each step carries percent_of_class and the 13D accession number as citation. Abstains (abstained=True) when the issuer has no verified control edge — never fabricates a chain from sub-50% or unclassified stakes.

Scope, precisely. Single-hop control reaches large caps: 20 of 825 controlled issuers carry >=$10B of institutional ownership and 97 carry >=$1B. It is the MULTI-HOP pyramids that are small-cap — those top out around $1.5B — so treat chain depth as a small-cap governance signal while single-hop control is general-purpose. Most large caps still have no >=50% holder and correctly abstain.

Each step carries institutional_value_usd (a 13F size proxy, null for the ~25% of issuers with no institutional coverage) so results can be ranked by materiality.

Examples: control_chain("TMUS") returns Deutsche Telekom -> T-Mobile US (74.3%, $95B). control_chain("Income Opportunity Realty") returns the 3-hop pyramid Basic Capital -> American Realty (62%) -> Transcontinental (83%) -> Income Opportunity (85%).

Name the director(s) who bridge two companies' boards.

Walks the persisted SHARES_DIRECTOR edge, returning the alternating company chain and the bridging director(s) per hop. Abstains when no shared-director path exists within max_hops.

Read the bridging director, not the path's existence. In this dataset every well-connected pair of companies is linked within 4 hops (measured), so "are these boards connected?" is essentially always yes and carries no information. The useful output is who the named connector is, and how short the link is.

Example: board_interlock_path("AAPL", "JPM") -> AAPL — JPM via BELL JAMES A (1 hop).

Find the de-facto activist coalition (wolf-pack) around a named 13D filer.

Two activists are linked when they co-target >=min_shared_targets of the same issuers via 13D. Custodial/broker hubs are scrubbed before the connected component is formed (precision discipline), so the coalition reflects coordinated activists, not shared custodians. Returns members + coalition diameter. Abstains when the filer sits in no multi-member coalition (or was scrubbed as a custodial hub).

Example: activist_coalition("ICAHN CARL C") -> the ~22-member scrubbed cluster (Bulldog/Goldstein, GAMCO/Gabelli, Karpus, Icahn, Dolan family).

Supporting ownership context for one issuer: top holders, board size, control status.

Grounds the headline chain/path/coalition answers with the surrounding facts — the largest 13D/13G beneficial owners (with percent_of_class), director/officer counts, and whether any verified >=50% control edge exists. Abstains only if the company can't be resolved.

Return the curated secgraph ownership schema and tool catalog for grounding.

A hand-curated summary (nodes, relationship semantics, temporal-trust caveats, and the curated tool catalog) — not a raw graph dump — so an agent can pick the right tool. Includes the honest limits (CIK-only, snapshot vs dated layers, no prediction).