# footnote-mcp (oci · ghcr.io/kazkozdev/footnote-mcp:0.2.3)

Source-grounded web research: search, extraction, verification, and browser automation.

- Trust score: 36/100 (low)
- Change this week: 0
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-08-03

## Components

- oci · `ghcr.io/kazkozdev/footnote-mcp:0.2.3`: 36/100 (this document), [markdown](https://verifymcp.io/servers/kazkozdev-footnote-mcp/ghcr-io-kazkozdev-footnote-mcp-0-2-3.md), [page](https://verifymcp.io/servers/kazkozdev-footnote-mcp/ghcr-io-kazkozdev-footnote-mcp-0-2-3)
- pypi · `footnote-mcp`: 36/100, [markdown](https://verifymcp.io/servers/kazkozdev-footnote-mcp/footnote-mcp.md), [page](https://verifymcp.io/servers/kazkozdev-footnote-mcp/footnote-mcp)

## Channel facts

- Registry: `oci`
- Package: `ghcr.io/kazkozdev/footnote-mcp:0.2.3`
- Transport: `stdio`

## 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 2026-08-03.

- **Supply Chain Security**: 0/100
  - Malware scan not yet available for this package.
  - CVE data not yet available for this package.
  - Install-script risk not yet assessed.
  - Dependency-health data not yet available.
- **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 (MIT).
  - Actively maintained (last published 10 days ago).
  - Disclosure check failed: no security disclosure policy was found in the source repository.
- **Schema Quality & AI Usability**: 54/100
  - AI-judged instruction clarity (fair).
  - Tool/resource definitions use about 2733 tokens (~60/item across 45 items; 45 tools + 0 resources), lean.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 23/100
  - Stability observed for 7 of 30 days with no destabilising changes; credit accrues until the full window elapses.
- **Tool Coverage**: 77/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 32% of tool parameters carry a description.
- **Capabilities**: 100/100
  - Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.

**Unverified: 1 category.** A category scored 0 because we could not verify it: 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

### Claude

```bash
claude mcp add kazkozdev-footnote-mcp -- docker run --rm -i ghcr.io/kazkozdev/footnote-mcp:0.2.3
```

### Codex

```bash
codex mcp add kazkozdev-footnote-mcp -- docker run --rm -i ghcr.io/kazkozdev/footnote-mcp:0.2.3
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "kazkozdev-footnote-mcp": {
      "type": "local",
      "command": [
        "docker",
        "run",
        "--rm",
        "-i",
        "ghcr.io/kazkozdev/footnote-mcp:0.2.3"
      ],
      "enabled": true
    }
  }
}
```

### Hermes

```yaml
mcp_servers:
  kazkozdev-footnote-mcp:
    command: "docker"
    args: ["run", "--rm", "-i", "ghcr.io/kazkozdev/footnote-mcp:0.2.3"]
```

### Other

```json
{
  "mcpServers": {
    "kazkozdev-footnote-mcp": {
      "command": "docker",
      "args": [
        "run",
        "--rm",
        "-i",
        "ghcr.io/kazkozdev/footnote-mcp:0.2.3"
      ]
    }
  }
}
```

## 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.

### 2026-08-03 (score 36, +3)

- [functional improvement] Stability: unverified → 0.23

### 2026-07-31 (score 33, −3)

- [functional] We updated how we score, so this day's move reflects our rubric, not a change to the server

### 2026-07-27 (score 36, +25)

- [functional improvement] Tool coverage: unverified → 100
- [functional] First check of Tool coverage: 32
- [functional] First check of Schema quality: fair
- [functional] First check of Schema quality: fail
- [functional] First check of Schema quality: pass

### 2026-07-26 (score 11)

First indexed and scored.

- [functional] We updated how we score, so this day's move reflects our rubric, not a change to the server

## MCP tools (45)

### `web_search` (~119 tokens)

General web discovery. Uses configured SearXNG first, then keyed providers, then scraped Bing + DuckDuckGo. Returns normalized titles, URLs, snippets and scores.

Input parameters:

- `lang` (string): Language code: en, ru, etc.
- `num` (integer): Max results to return
- `provider` (string): auto | searxng | tavily | brave | google | scrape
- `query` (string, required): Search query
- `semantic` (boolean): Rerank results by meaning using local bge-m3 embeddings

### `web_deep_search` (~127 tokens)

Intent-routed deep research: discover through web/papers/encyclopedia/GitHub/archive sources, then fetch, extract, rerank, and return source-grounded context.

Input parameters:

- `lang` (string): Language code
- `num` (integer): Maximum discovery results before fetching
- `provider` (string): Web provider: auto | searxng | tavily | brave | google | scrape
- `query` (string, required): Search query
- `sources` (array): Optional explicit source intents: web, papers, encyclopedia, github, archive. Empty enables automatic routing.

### `web_read` (~53 tokens)

Fetch and extract text from a single URL. Returns extracted content (markdown).

Input parameters:

- `lang` (string): Language code
- `url` (string, required): Full URL to fetch and extract
- `use_cache` (boolean)

### `web_extract_tables` (~66 tokens)

Fetch a page, parse HTML tables, and return structured columns/rows with source URL provenance.

Input parameters:

- `lang` (string)
- `max_rows` (integer)
- `max_tables` (integer)
- `url` (string, required)
- `use_cache` (boolean)

### `web_detect_downloads` (~50 tokens)

Detect downloadable CSV/XLS/XLSX/PDF/JSON/XML links on a page.

Input parameters:

- `lang` (string)
- `max_links` (integer)
- `url` (string, required)

### `web_parse_file` (~64 tokens)

Download and parse CSV/TSV/XLSX/PDF/JSON files, returning structured rows or extracted text with provenance.

Input parameters:

- `lang` (string)
- `max_rows` (integer)
- `url` (string, required)
- `use_cache` (boolean)

### `web_fetch_json` (~57 tokens)

Fetch an API/JSON URL directly and return parsed JSON with source URL provenance and persistent cache.

Input parameters:

- `lang` (string)
- `timeout` (integer)
- `url` (string, required)
- `use_cache` (boolean)

### `check_date_completeness` (~108 tokens)

Validate date-range completeness for structured results. Supports day, week, and month granularity.

Input parameters:

- `actual_items` (array, required)
- `calendar` (string): calendar | business_day | crypto_24_7 | forex_weekday | us_business_day | ru_business_day
- `end_date` (string, required): YYYY-MM-DD
- `granularity` (string): day | week | month
- `holidays` (array)
- `start_date` (string, required): YYYY-MM-DD

### `classify_source` (~59 tokens)

Classify a source as official, aggregator, blog, forum, interactive, blocked, or error.

Input parameters:

- `content_type` (string)
- `status_code` (integer)
- `text_sample` (string)
- `url` (string, required)

### `generate_search_queries` (~52 tokens)

Generate specialized search queries using operators like site:, filetype:, API, CSV, and data-table variants.

Input parameters:

- `max_queries` (integer)
- `requirements` (object)
- `task` (string, required)

### `resolve_units` (~32 tokens)

Resolve units, currencies, and currency pairs from text so incompatible rows can be rejected.

Input parameters:

- `text` (string, required)

### `validate_unit_rows` (~51 tokens)

Reject structured rows that are incompatible with the expected unit, currency, or currency pair.

Input parameters:

- `expected_unit_or_pair` (string, required)
- `rows` (array, required)
- `text_fields` (array)

### `evidence_entailment` (~75 tokens)

Strict entailment check for claim vs source excerpt. Supports heuristic, Ollama judge, or auto fallback.

Input parameters:

- `backend` (string): auto | heuristic | ollama | local_nli
- `claim` (string, required)
- `model` (string): Optional Ollama model name
- `source_excerpt` (string, required)

### `tool_spec_propose` (~58 tokens)

Propose a controlled task-specific extraction recipe spec when generic tools cannot extract structured rows.

Input parameters:

- `desired_output` (string)
- `observed_failure` (string)
- `source_url` (string)
- `task` (string, required)

### `tool_code_generate` (~37 tokens)

Generate a safe starter extraction recipe function for the proposed spec. The result must still be validated before running.

Input parameters:

- `spec` (object)

### `tool_code_validate` (~40 tokens)

Statically validate task-specific extraction code against the safe recipe contract and allowlist.

Input parameters:

- `code` (string, required)
- `max_chars` (integer)

### `tool_code_run_sandboxed` (~72 tokens)

Run validated extraction recipe code in a separate limited subprocess. Code must define extract(source_text, input_payload).

Input parameters:

- `code` (string, required)
- `input_payload` (object)
- `max_output_chars` (integer)
- `source_text` (string)
- `timeout` (integer)

### `tool_promote` (~77 tokens)

Save a validated extraction recipe as reusable memory after an optional smoke test. Does not edit the MCP server.

Input parameters:

- `code` (string, required)
- `expected_min_rows` (integer)
- `input_payload` (object)
- `name` (string, required)
- `sample_source_text` (string)
- `spec` (object)

### `source_cache_get` (~25 tokens)

Read persistent source cache entry for a URL.

Input parameters:

- `url` (string, required)

### `source_cache_put` (~33 tokens)

Write arbitrary parsed source payload into persistent source cache.

Input parameters:

- `payload` (object, required)
- `url` (string, required)

### `build_research_debug_report` (~68 tokens)

Build a compact diagnostic report for a research run: queries, URLs, source quality, and verification gaps.

Input parameters:

- `requirements` (object)
- `search_memory` (object)
- `sources` (array)
- `task` (string, required)
- `verification` (object)

### `startup_health_check` (~25 tokens)

Check optional parser, OCR, browser, and cache dependencies for this MCP server.

### `web_navigate` (~28 tokens)

Navigate the browser to a URL.

Input parameters:

- `url` (string, required): Full URL to navigate to

### `web_snapshot` (~42 tokens)

Capture the current page state: URL, title, accessibility tree with stable refs (@e1, @e2...), and visible text. Use this before clicking/typing.

### `web_click` (~43 tokens)

Click an interactive element by its ref (e.g. @e3 from web_snapshot).

Input parameters:

- `ref` (string, required): Element ref from web_snapshot, like @e3

### `web_type` (~52 tokens)

Type text into an input field by its ref.

Input parameters:

- `ref` (string, required): Input field ref from web_snapshot
- `submit` (boolean): Press Enter after typing
- `text` (string, required): Text to type

### `web_extract` (~58 tokens)

Extract text from the page. refs: comma-separated @eN, or 'visible' for all visible text, or 'all' for full HTML.

Input parameters:

- `refs` (string, required): Comma-separated refs, or 'visible', or 'all'

### `web_scroll` (~34 tokens)

Scroll the page: up, down, top, bottom.

Input parameters:

- `direction` (string, required): up | down | top | bottom

### `browser_extract_tables` (~38 tokens)

Extract visible tables from the current browser page after navigation/interactions.

Input parameters:

- `max_rows` (integer)
- `max_tables` (integer)

### `browser_set_date_range` (~53 tokens)

Best-effort browser date-range setter for interactive pages with date inputs.

Input parameters:

- `end_date` (string, required): YYYY-MM-DD
- `start_date` (string, required): YYYY-MM-DD
- `submit` (boolean)

### `browser_extract_tables_for_date_range` (~67 tokens)

Set date range in the current browser page, submit it, then extract visible tables.

Input parameters:

- `end_date` (string, required): YYYY-MM-DD
- `max_rows` (integer)
- `max_tables` (integer)
- `start_date` (string, required): YYYY-MM-DD

### `web_archive_fetch` (~73 tokens)

Find the closest Wayback Machine snapshot of a URL (for dead or changed sources) and optionally read its text.

Input parameters:

- `fetch_text` (boolean)
- `lang` (string)
- `timestamp` (string): Optional target YYYYMMDD or YYYYMMDDhhmmss
- `url` (string, required)

### `papers_search` (~66 tokens)

Search scientific publications through Crossref and arXiv. Uses no API key and returns one normalized paper result shape.

Input parameters:

- `lang` (string)
- `num` (integer)
- `query` (string, required)
- `source` (string): auto | crossref | arxiv

### `encyclopedia_search` (~89 tokens)

Search Wikipedia articles and Wikidata entities, or execute a read-only Wikidata SPARQL query.

Input parameters:

- `lang` (string)
- `num` (integer)
- `query` (string, required): Entity/topic query; still required when sparql is supplied
- `source` (string): auto | wikipedia | wikidata
- `sparql` (string): Optional read-only SPARQL SELECT query

### `github_search` (~64 tokens)

Search public GitHub repositories, issues, code, or commits. Authentication is optional; zero-key requests use GitHub's public rate limit.

Input parameters:

- `kind` (string): repositories | issues | code | commits
- `num` (integer)
- `query` (string, required)

### `archive_search` (~105 tokens)

Find archived captures of a URL or host through the Wayback Machine and Common Crawl; optionally extract archived text.

Input parameters:

- `fetch_text` (boolean)
- `lang` (string)
- `num` (integer)
- `source` (string): auto | wayback | common_crawl
- `timestamp` (string): Optional Wayback target YYYYMMDD or YYYYMMDDhhmmss
- `url` (string, required): Full URL, host, or Common Crawl URL pattern

### `web_search_recent` (~65 tokens)

Web search restricted to a recency window via DuckDuckGo's date filter (day/week/month/year).

Input parameters:

- `freshness` (string): day | week | month | year
- `lang` (string)
- `num` (integer)
- `query` (string, required)

### `corroborate_claim` (~67 tokens)

Triangulate a claim across multiple source excerpts; returns a corroboration verdict (corroborated/conflicting/single_source/...).

Input parameters:

- `backend` (string): heuristic | auto | ollama | local_nli
- `claim` (string, required)
- `excerpts` (array, required)

### `locate_claim_span` (~59 tokens)

Locate the sentence(s) in a source that best support a claim, with character offsets and a containment score (span-level provenance).

Input parameters:

- `claim` (string, required)
- `max_spans` (integer)
- `source_text` (string, required)

### `recipe_registry` (~62 tokens)

Manage promoted extraction recipes: list, get, run, or delete saved recipes.

Input parameters:

- `action` (string, required): list | get | run | delete
- `input_payload` (object)
- `recipe_id` (string)
- `source_text` (string)

### `web_fetch_authenticated` (~68 tokens)

Fetch a page that needs cookies or custom headers (logged-in or gated pages).

Input parameters:

- `cookies` (object): name→value cookie map
- `headers` (object): extra request headers
- `lang` (string)
- `timeout` (integer)
- `url` (string, required)

### `web_crawl` (~71 tokens)

Breadth-first crawl from a start URL, fetching and extracting each page. Stays on the start host by default. Capped at 50 pages.

Input parameters:

- `lang` (string)
- `max_pages` (integer)
- `same_domain` (boolean)
- `start_url` (string, required)

### `export_dataset` (~71 tokens)

Write extracted rows to a consolidated file (csv | xlsx | json) and return the path.

Input parameters:

- `columns` (array)
- `format` (string): csv | xlsx | json
- `path` (string): Optional output path; defaults to cache exports dir
- `rows` (array, required)

### `reconcile_time_series` (~61 tokens)

Align several time series on a common key, compute deltas vs the first series, and flag missing keys and outliers.

Input parameters:

- `on` (string): Key field to align on
- `series` (array, required)
- `value_field` (string)

### `web_screenshot` (~49 tokens)

Capture a PNG screenshot of the current browser page, save it to disk, and optionally OCR text locked inside the image.

Input parameters:

- `full_page` (boolean)
- `ocr` (boolean)

## Diagnostics

Captured diagnostic sections: Provenance. The full working is on the page: https://verifymcp.io/servers/kazkozdev-footnote-mcp/ghcr-io-kazkozdev-footnote-mcp-0-2-3#diagnostics

## Score history

- 2026-08-03: 36
- 2026-08-02: 33
- 2026-08-01: 33
- 2026-07-31: 33
- 2026-07-30: 36
- 2026-07-29: 36
- 2026-07-28: 36
- 2026-07-27: 36
- 2026-07-26: 11

## Links

- Repository: https://github.com/KazKozDev/footnote-mcp
- Changelog RSS feed: https://verifymcp.io/servers/kazkozdev-footnote-mcp/ghcr-io-kazkozdev-footnote-mcp-0-2-3/changelog.xml
- Changelog JSON feed: https://verifymcp.io/servers/kazkozdev-footnote-mcp/ghcr-io-kazkozdev-footnote-mcp-0-2-3/changelog.json
- HTML version of this page: https://verifymcp.io/servers/kazkozdev-footnote-mcp/ghcr-io-kazkozdev-footnote-mcp-0-2-3