cbam-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| CBAM_MCP_DATA_DIR | No | Directory to read the JSON data files from instead of the bundled cbam_mcp/data/ directory (set when the data was refreshed elsewhere with `cbam-mcp refresh --data-dir DIR`). Optional; if unset, the bundled data is used. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| cbam_scopeA | Whether goods under a CN code are in the scope of the EU Carbon Border Adjustment Mechanism, from Annex I of Regulation (EU) 2023/956 as amended by Regulation (EU) 2025/2083 (consolidated text of 2025-10-20). Returns status 'in_scope', 'partially_in_scope' or 'not_in_scope' with the reason; the deciding Annex I line (goods category, greenhouse gases, exceptions, 'ex' flag); whether Annex II limits the goods to direct emissions; the de minimis texts (Article 2a, Annex VII) where they apply; the CN 2026 description. 'partially_in_scope' means an 'ex' line (only part of the code is covered: read the line's text) or a heading whose sub-codes differ. For codes shorter than 8 digits it lists the CN 2026 sub-codes in and out of scope (up to |
| default_valueA | CBAM default values for goods under a CN code from one country of origin, from the Commission's Excel of definitive-period default values (the binding values are in Annex I to Implementing Regulation (EU) 2025/2621 as replaced by Implementing Regulation (EU) 2026/1740). Returns per matching table line: direct, indirect and total embedded emissions in tCO2e per tonne of good as listed, before the mark-up (markup_rule quotes the rule from the consolidated text; this tool does not apply it), the production-route letter and its meaning, the 'Other countries and territories' values where Annex I says to use them for a listed country (no line, or '–'), and the Annex IV value for precursors whose country of production is unknown. A country name the tool does not recognise is an error, never a fallback; for a third country the tables do not list, ask for country 'Other countries and territories'. EU Member States and Annex III origins (e.g. Norway, Switzerland) get an explanation instead of values; a code outside CBAM scope gets no value. No values for electricity (2716) and no indirect-emission factors: those are Annexes II and III, not in the Excel. Not legally binding. |
| compare_originsA | The CBAM default values of one CN code for several countries of origin side by side (up to 30), with the 'Other countries and territories' row and the Annex IV value, in tCO2e per tonne of good as listed in the Commission's Excel (not legally binding; binding: Implementing Regulation (EU) 2025/2621 as corrected by 2026/1740). Each country shows where its values come from, including the fallback rule when used. |
| cn_describeA | The Combined Nomenclature description of a code in CN 2026 or CN 2025: label, self-explanatory text, the hierarchy above it (section, chapter, heading, subheading), its sub-codes, and whether the other year differs. Covers chapters 25, 28, 31, 72, 73, 76 and headings 2601 and 2716 (where the CBAM goods are); other codes return found=false. Not legally binding: the binding CN is the Official Journal text of Commission Implementing Regulation (EU) 2025/1926 (CN 2026) or 2024/2522 (CN 2025). |
| sourcesA | Where the answers come from: for each bundled dataset the URL, version label, retrieval date, SHA-256, row counts, legal status, licence and attribution line, the result of the check against the Official Journal text, and the later amending or correcting acts found in CELLAR at the last refresh. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 5 tools
Each tool targets a distinct output: scope determination (cbam_scope), values for one origin (default_value), values across many origins (compare_origins), nomenclature lookup (cn_describe), and provenance (sources). The only real overlap is default_value vs. compare_origins, which are effectively single-origin vs. batched multi-origin views of the same data, but both descriptions make the split clear.
Names are all snake_case and readable, but verb placement is inconsistent: verb_noun (compare_origins), prefixed noun (cbam_scope), bare noun (default_value, sources), and noun_verb (cn_describe). A predictable pattern is not established, though an agent can still infer each tool's role.
Five tools are well-scoped for a reference-data lookup service: scope, single-origin values, multi-origin comparison, nomenclature description, and source metadata. No tool feels redundant or filler, and nothing essential appears missing from the count perspective.
The surface covers the core CBAM data workflows (scope check, default values, comparison, code descriptions, provenance). Gaps are minor-to-moderate: no keyword/reverse lookup for CN codes, no listing of in-scope goods, and markup is returned but never applied in a calculation, though that may be intentionally out of scope.