# Postflop GTO Solver (mcpb · postflop-mcp.mcpb)

Solve heads-up postflop poker spots with a GTO solver and query the solution

- Trust score: 36/100 (low)
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-11-21

## Components

- mcpb · `postflop-mcp.mcpb`: 36/100 (this document), [markdown](https://verifymcp.io/servers/sawadakaku-postflop-mcp/postflop-mcp.md), [page](https://verifymcp.io/servers/sawadakaku-postflop-mcp/postflop-mcp)

## Channel facts

- Registry: `mcpb`
- Package: `https://github.com/sawadakaku/postflop-mcp/releases/download/v0.1.0/postflop-mcp.mcpb`
- Transport: `stdio`
- SHA-256: `3650c6d6d4f5aa494b57ea91e9b3e6fa158dea8077f9fce958765eca6416bee0`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-11-21.

- **Supply Chain Security**: 0/100
  - No malware scan is available for this kind of package: the supply-chain vendors we use do not cover it. This is a permanent gap in our coverage, not a finding about the package.
  - Known CVEs could not be checked: we found no dependency list to check in this bundle: either it runs a compiled binary, whose bundled manifests we do not read, or none of the package.json, requirements.txt, pyproject.toml or uv.lock files it ships for its runtime declares a dependency we can read (a requirements.txt that only points at a setup.py project, a local path, or a nested requirements file we cannot follow counts as none).
  - Install-script risk not yet assessed.
  - Dependency health could not be checked: we found no dependency list to check in this bundle: either it runs a compiled binary, whose bundled manifests we do not read, or none of the package.json, requirements.txt, pyproject.toml or uv.lock files it ships for its runtime declares a dependency we can read (a requirements.txt that only points at a setup.py project, a local path, or a nested requirements file we cannot follow counts as none).
- **Provenance & Transparency**: 45/100
  - Source repository is publicly reachable at the declared URL.
  - Provenance check failed: no build-provenance attestation is published.
  - Clear OSI-approved license (AGPL-3.0-or-later).
  - Actively maintained (last published 4 days ago).
  - Disclosure check failed: no security disclosure policy was found in the source repository.
- **Schema Quality & AI Usability**: 58/100
  - AI-judged instruction clarity (good).
  - Context-footprint check failed: tool/resource definitions use about 3158 tokens (~225/item across 14 items; 14 tools + 0 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 0/100
  - Stability not yet verified: not enough scan history yet (needs a 30-day window).
- **Tool Coverage**: 99/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 96% of tool parameters carry a description.
- **Tool Safety**: 100/100
  - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.
  - We read all 14 captured tool definition(s), and no name or description among them implies an irreversible operation.
  - An AI judge read all 15 captured unit(s) of tool text and found none that tries to manipulate the model reading it.
- **Capabilities**: 60/100
  - Spec-recency check failed: implements MCP spec 2025-06-18; the latest is 2026-07-28.

**Unverified: 2 categories.** Categories scored 0 because we could not verify them: a data source with nothing on this package, evidence we could not reach, or a check we could not run. We only credit what we can confirm.

## Install

### How do I install the Postflop GTO Solver MCP server?

Postflop GTO Solver ships as an MCPB bundle, a single file you download and open in an MCP host that supports MCPB bundles, such as Claude Desktop. The host reads the launch command from the bundle's own manifest, so there is no command to copy. Before opening it, check the file's SHA-256 matches the one the MCP registry declares for it, 3650c6d6d4f5aa494b57ea91e9b3e6fa158dea8077f9fce958765eca6416bee0; the commands to do that are on this page. If it does not match, do not open it.

- Download bundle: `https://github.com/sawadakaku/postflop-mcp/releases/download/v0.1.0/postflop-mcp.mcpb`
- SHA-256: `3650c6d6d4f5aa494b57ea91e9b3e6fa158dea8077f9fce958765eca6416bee0`

### Verify on macOS / Linux

```bash
echo "3650c6d6d4f5aa494b57ea91e9b3e6fa158dea8077f9fce958765eca6416bee0  postflop-mcp.mcpb" | shasum -a 256 -c
```

### Verify on Windows (PowerShell)

```powershell
(Get-FileHash -Algorithm SHA256 .\postflop-mcp.mcpb).Hash -eq "3650c6d6d4f5aa494b57ea91e9b3e6fa158dea8077f9fce958765eca6416bee0"
```

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-11-17 (score 36)

First indexed and scored.

## MCP tools (14)

### `solve` (~473 tokens)

Solve a spot (call this first; solving another spot replaces it).
If the same spot was solved before, it loads from the cache in about a second. Otherwise it is solved and saved. From the river it takes under a second, from the turn seconds to tens of seconds, from the flop tens of seconds to minutes.
Returns: the spot, the player names, the root node (same as node_info), the accuracy (exploitability_bb) and the memory used.
With dry_run=true it only returns the memory needed, without solving.

Input parameters:

- `board` (string, required): 3 to 5 board cards, e.g. "QcQdJc" or "Qc Qd Jc 2d". The number of cards sets the starting street
- `dry_run` (boolean): If true, only return the memory needed (MB) without solving
- `ip_range` (string, required): Range of IP (the player who acts last)
- `max_iterations` (integer): Maximum iterations (default 1000)
- `names` (object): Player names, e.g. {"oop": "BB", "ip": "BTN"}. Defaults to OOP / IP
- `oop_range` (string, required): Range of OOP (the player who acts first), e.g. "AA-TT,AKs,A5s-A2s,KQo:0.5" (:0.5 is a weight; write ranges high to low: 99-22, AJo-A8o)
- `pot_bb` (number, required): Pot at the start of the street, in bb
- `stack_bb` (number, required): Effective stack at the start of the street, in bb (what is left behind)
- `target_exploitability_pct` (number): Target accuracy as % of the pot (default 0.3). Smaller is more accurate and slower
- `tree` (object): Bet sizes per street as [OOP, IP] 'bet/raise' strings, e.g. {"turn": ["50%,75%/a", "33%,75%/a"]}. % is relative to the pot, a is all-in, 2.5x is a raise multiplier, and an empty "/" means check/call…

### `node_info` (~122 tokens)

Who acts at this node, what they can do, and how the whole range plays.
Returns: player to act, pot, stacks, amount to call, and per action the frequency (freq), combos and average EV (over the hands that take it), plus both players' overall equity. At a chance node, the cards that can come.

Input parameters:

- `path` (string): Node path, e.g. `x b75 c / Js / x`. Empty for the root of the spot. May start with `@sub3: ` or `#A `

### `hand_detail` (~170 tokens)

Frequency, EV and equity of a hand for each action.
hand is a combo (Ah8s) or a class (A8o / A8s / 88). A class lists each suit combination, which helps compare blockers.
Without player, the player to act. For the player not to act, only EV and equity.

Input parameters:

- `hand` (string, required): A combo (Ah8s) or a class (A8o / A8s / 88)
- `path` (string): Node path, e.g. `x b75 c / Js / x`. Empty for the root of the spot. May start with `@sub3: ` or `#A `
- `player` (string): Player: a name given in solve's names, or OOP / IP

### `range` (~118 tokens)

A player's range at this node: share and equity by hand class (top 20) and by made-hand label (full house, set, top pair, overpair, two pair on board (A kicker), ...).

Input parameters:

- `path` (string): Node path, e.g. `x b75 c / Js / x`. Empty for the root of the spot. May start with `@sub3: ` or `#A `
- `player` (string, required): Player: a name given in solve's names, or OOP / IP

### `action_range` (~138 tokens)

The hands with which the player to act takes action (x / c / f / allin / b75, ...).
Returns: the action's frequency, the breakdown by class and label, and the value/bluff split (value = equity of 50% or more against the opponent's whole range at this node).

Input parameters:

- `action` (string, required): Action (x / c / f / allin / b75, ...)
- `path` (string): Node path, e.g. `x b75 c / Js / x`. Empty for the root of the spot. May start with `@sub3: ` or `#A `

### `lookahead` (~101 tokens)

Frequencies for the next few actions and the top classes taking each one. Stops at chance nodes; add a card to the path to look further.

Input parameters:

- `depth` (integer): How many actions ahead (1-4, default 2)
- `path` (string): Node path, e.g. `x b75 c / Js / x`. Empty for the root of the spot. May start with `@sub3: ` or `#A `

### `range_chart` (~149 tokens)

Save a player's range as a 13x13 hand chart (HTML).
Use this to show a range instead of writing it out as a table. Show the user the returned file (send it if you can, otherwise give the path).
For the player to act, each cell is colored by action frequencies and shows EVs when clicked. Otherwise cells are colored by equity.

Input parameters:

- `path` (string): Node path, e.g. `x b75 c / Js / x`. Empty for the root of the spot. May start with `@sub3: ` or `#A `
- `player` (string, required): Player: a name given in solve's names, or OOP / IP

### `tree_view` (~163 tokens)

Save the next few actions from this node as a clickable game tree (HTML).
Clicking a node shows the player to act, the pot, frequencies and EVs per action, and both players' 13x13 range charts. Use it for "what happens next?" or "show me this line", and show the user the returned file.
Stops at chance nodes and terminal nodes. Call it again with a card added to the path to go further.

Input parameters:

- `depth` (integer): How many actions ahead (1-4, default 3)
- `path` (string): Node path, e.g. `x b75 c / Js / x`. Empty for the root of the spot. May start with `@sub3: ` or `#A `

### `suit_chart` (~142 tokens)

Save a table (HTML) of every suit combination of a hand class (A8o / A8s / 88) of the player to act.
Shows each combo's frequencies, EV and equity, and how much of the opponent's value and bluffs it blocks. Use it to explain blockers, and show the user the returned file.

Input parameters:

- `hand` (string, required): Hand class (A8o / A8s / 88)
- `path` (string): Node path, e.g. `x b75 c / Js / x`. Empty for the root of the spot. May start with `@sub3: ` or `#A `

### `resolve_from` (~382 tokens)

Re-solve from the start of this node's street with changes (branches outside the tree).
The start of the last street in path (right after its last card) becomes the root of the subgame. Earlier streets keep the original solution.
Prefix a path with the returned sub (`@sub3: x b50`) to query the subgame with any tool. Always state the premise in your answer. From the river it takes under a second, from the turn seconds, from the flop tens of seconds to minutes.

Input parameters:

- `locks` (array): Strategy locks, e.g. [{"hands": "88-33", "strategy": {"allin": 1}}]. Without path, the node given in path; with path, a path from the subgame root. Only the given hands are locked; the solver plays t…
- `path` (string, required): Node path, e.g. `x b75 c / Js / x`. Empty for the root of the spot. May start with `@sub3: ` or `#A `
- `ranges` (array): Range changes, e.g. [{"player": "IP", "hands": "AK,AQ", "scale": 0}] (multiplies the reach; 0 removes the hands)
- `stack_bb` (number): Effective stack at the subgame root, in bb. To compare stack depths, compare_stacks is quicker
- `tree` (object): Bet sizes per street as [OOP, IP] 'bet/raise' strings, e.g. {"turn": ["50%,75%/a", "33%,75%/a"]}. % is relative to the pot, a is all-in, 2.5x is a raise multiplier, and an empty "/" means check/call…

### `exploit` (~301 tokens)

How to play if the opponent is biased. path is a node where the opponent acts (turn or river).
Turn words like "bluffs too much" into concrete hands and strategy first (e.g. hands="K9s-K5s", strategy={"allin": 1}).
The opponent keeps the original strategy except that hands play strategy; your best response is re-solved.
Returns: ev_bb (GTO vs GTO / GTO vs biased / exploit vs biased / exploit vs GTO), gain_bb (what exploiting gains), risk_bb (what it loses if the opponent was not biased), hero_node (your frequencies after the opponent's then), assumption, notes and premise. Always state the assumption and premise.

Input parameters:

- `hand` (string): Your hand (Ah8s / A8o). Also returns how it plays in GTO and in the exploit
- `hands` (string, required): Opponent hands to bias (88-33 / A5s+ / KQo, ...)
- `path` (string, required): Node path, e.g. `x b75 c / Js / x`. Empty for the root of the spot. May start with `@sub3: ` or `#A `
- `strategy` (object, required): Frequencies for those hands, e.g. {"allin": 1}
- `then` (string): The opponent action after which to show your node (default: the heaviest action in strategy)

### `bookmark` (~126 tokens)

Save, list or delete named branches.
name and path save, name and delete=true delete, no arguments lists. Names have no spaces (e.g. A, line1).
A saved branch expands when a path starts with #name (e.g. `#A x b75`).

Input parameters:

- `delete` (boolean)
- `name` (string): Bookmark name
- `path` (string): Node path, e.g. `x b75 c / Js / x`. Empty for the root of the spot. May start with `@sub3: ` or `#A `

### `compare_lines` (~130 tokens)

Compare two or more lines (paths, @sub3: ..., #A) side by side.
For each line: player to act, pot, frequency and EV per action, and both players' overall equity. With hand (Ah8s / A8o), also that hand's frequencies, EV and equity (without player, the player to act in each line).

Input parameters:

- `hand` (string): Hand to compare (Ah8s / A8o)
- `lines` (array, required)
- `player` (string): Player: a name given in solve's names, or OOP / IP

### `compare_stacks` (~242 tokens)

Re-solve the same node with only the stack depth changed and compare. Also saves a comparison view (HTML).
Re-solves from the start of path's street (from the root of the spot with from_root=true). The first column is always the original solution.
Always state the returned premise (where the stack was changed, and that earlier play used the original stack).
With shallow stacks large bets merge into all-in, so some columns may not follow the path (error).

Input parameters:

- `from_root` (boolean): If true, re-solve from the root of the spot (slower)
- `hand` (string): Hand to include (Ah8s / A8o)
- `path` (string, required): Node path, e.g. `x b75 c / Js / x`. Empty for the root of the spot. May start with `@sub3: ` or `#A `
- `player` (string): Player: a name given in solve's names, or OOP / IP
- `stacks` (array): Effective stacks at the re-solve root, in bb. Defaults to SPR 1, 2 and 4

## Diagnostics

Captured diagnostic sections: Provenance. The full working is on the page: https://verifymcp.io/servers/sawadakaku-postflop-mcp/postflop-mcp#diagnostics

## Score history

- 2026-11-21: 36
- 2026-11-20: 36
- 2026-11-19: 36
- 2026-11-18: 36
- 2026-11-17: 36

## Common questions

### What is the Postflop GTO Solver MCP server?

Postflop GTO Solver is an MCP server listed in the public MCP registry as io.github.sawadakaku/postflop-mcp. Solve heads-up postflop poker spots with a GTO solver and query the solution. This page covers its MCPB bundle (https://github.com/sawadakaku/postflop-mcp/releases/download/v0.1.0/postflop-mcp.mcpb).

### Is the Postflop GTO Solver MCP server safe to use?

Postflop GTO Solver scores 36 out of 100 on VerifyMCP. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.

### What tools does the Postflop GTO Solver MCP server expose?

Postflop GTO Solver exposes 14 tools: solve, node_info, hand_detail, range, action_range, and 9 more. Their descriptions and schemas cost roughly 2,757 tokens of context every time the server is loaded.

### What licence is the Postflop GTO Solver MCP server under?

Postflop GTO Solver declares the AGPL-3.0-or-later licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.

### What did VerifyMCP test on the Postflop GTO Solver MCP server?

VerifyMCP ran 11 checks on Postflop GTO Solver, covering Source repository, Provenance, License, Maintenance and 7 more. 5 items could not be completed, a gap on our side rather than a finding about Postflop GTO Solver: Malware scan, Known CVEs, Install scripts, Dependency health and 1 more. 2 categories score 0 for that reason, because we only credit what we can confirm. Endpoint Security and Transport & Reachability do not apply to an MCPB bundle, so they were not tested. The score does not measure everything, and the scoring documentation lists what it leaves out.

## Links

- Repository: https://github.com/sawadakaku/postflop-mcp
- Changelog RSS feed: https://verifymcp.io/servers/sawadakaku-postflop-mcp/postflop-mcp.xml
- Changelog JSON feed: https://verifymcp.io/servers/sawadakaku-postflop-mcp/postflop-mcp/changelog.json
- Trust report (JSON): https://verifymcp.io/servers/sawadakaku-postflop-mcp/postflop-mcp.json
- HTML version of this page: https://verifymcp.io/servers/sawadakaku-postflop-mcp/postflop-mcp
