ai.geodesiclabs/governance-platform

Est-ce que ai.geodesiclabs/governance-platform est sûr ?

Au dernier audit MCP Hub, le serveur MCP « ai.geodesiclabs/governance-platform » est classé « Non encore testé ». Référencé dans l'annuaire. Seul le pentest comportemental fait foi, et il n'a pas encore été exécuté sur ce serveur. Le verdict est réévalué en continu ; consultez la fiche pour l'état à jour.

Méthode : MCP Hub analyse la description de chaque outil (couche 1) et, pour le verdict fort, exécute le serveur dans une micro-VM isolée avec leurres et réseau bloqué (couche 2). On publie des noms et des compteurs, jamais le contenu brut renvoyé par un outil.

Les 37 outils

• account_status≈ 125 tokens

  This account's plan, key usage, Blueprint counts, and the deployed platform build fingerprint (version, build, deployed).

  Paramètres, retour & exemple (1)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)

  Exemple d'appel

  {"name": "account_status", "arguments": {"api_key": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• analyze_anomaly≈ 225 tokens

  Explain whether a record fits the usual pattern for records like it, and which fields stand out. No Blueprint required.

  Paramètres, retour & exemple (3)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • blueprint · string · optionnel
    Discovery namespace used by discover_patterns
  • structured_data · object · requis
    The document's extracted fields as key/value pairs. Keys are open by design - your Blueprint's rules define what is checked

  Exemple d'appel

  {"name": "analyze_anomaly", "arguments": {"api_key": "texte", "structured_data": {}}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• approve_rule≈ 176 tokens

  Promote a rule discovered by discover_patterns into Blueprint-ready form.

  Paramètres, retour & exemple (3)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • rule_id · string · requis
    Discovered rule ID from discover_patterns
  • blueprint · string · optionnel
    Blueprint name (workflow_name) to use

  Exemple d'appel

  {"name": "approve_rule", "arguments": {"api_key": "texte", "rule_id": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• authorize_execution≈ 274 tokens

  Go/no-go for a real-world action (payment, filing, API write): runs full validation, then the Blueprint's execution gate. authorized=true only on PASS; REVIEW means do not proceed automatically. Different from validate: validate asks is this data correct, authorize_execution asks should this action happen.

  Paramètres, retour & exemple (3)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • blueprint · string · requis
    Blueprint name (workflow_name) to use
  • structured_data · object · requis
    The document's extracted fields as key/value pairs. Keys are open by design - your Blueprint's rules define what is checked

  Exemple d'appel

  {"name": "authorize_execution", "arguments": {"api_key": "texte", "blueprint": "texte", "structured_data": {}}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• check_blueprint_health≈ 242 tokens

  Static pre-deploy analysis of a Blueprint's rule set. Returns a health verdict - healthy, acceptable, fragile, rigid, split, brittle_islands, or unsatisfiable - with advice, including joint conflicts pairwise checks miss.

  Paramètres, retour & exemple (3)

  Paramètres

  • config · object · optionnel
    Raw blueprint config with derivation_rules and formal_constraints (used when 'blueprint' is not given)
  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • blueprint · string · optionnel
    Blueprint name (workflow_name) to use

  Exemple d'appel

  {"name": "check_blueprint_health", "arguments": {"api_key": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• check_drift≈ 226 tokens

  Check whether recent submissions still match the established pattern for this Blueprint. Returns a stability verdict and observation count.

  Paramètres, retour & exemple (3)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • blueprint · string · optionnel
    Blueprint name (workflow_name) to use
  • structured_data · object · requis
    The document's extracted fields as key/value pairs. Keys are open by design - your Blueprint's rules define what is checked

  Exemple d'appel

  {"name": "check_drift", "arguments": {"api_key": "texte", "structured_data": {}}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• check_realization≈ 255 tokens

  Structural realization analysis of a payload against the Blueprint's reference configuration (requires a 'realization' block; otherwise status=skipped). Diagnostics-tier tool; prefer validate or analyze_anomaly for standard checks.

  Paramètres, retour & exemple (3)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • blueprint · string · optionnel
    Blueprint name (workflow_name) to use
  • structured_data · object · requis
    The document's extracted fields as key/value pairs. Keys are open by design - your Blueprint's rules define what is checked

  Exemple d'appel

  {"name": "check_realization", "arguments": {"api_key": "texte", "structured_data": {}}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• compare_semantic_equivalence≈ 448 tokens

  Compare two payloads under the dual-hash design: content_hash is content_hash normalizes field order and numeric formatting. Semantic comparison preserves field roles; renaming requires an explicit bijection. With supplied rules, scalar types and whitespace remain significant. Structural similarity alone does not establish decision equivalence.

  Paramètres, retour & exemple (8)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • rules_a · array · optionnel
    Derivation rules for A
  • rules_b · array · optionnel
    Derivation rules for B
  • payload_a · object · requis
    First structured payload (arbitrary JSON object)
  • payload_b · object · requis
    Second structured payload to compare against payload_a
  • constraints_a · array · optionnel
    Formal constraints for A
  • constraints_b · array · optionnel
    Formal constraints for B
  • field_mapping · object · optionnel
    Explicit one-to-one field renaming from A to B

  Exemple d'appel

  {"name": "compare_semantic_equivalence", "arguments": {"api_key": "texte", "payload_a": {}, "payload_b": {}}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• counterfactual≈ 294 tokens

  Run the same data under two rule sets and compare which future states remain valid - what-if analysis for rule changes.

  Paramètres, retour & exemple (5)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • rules_b · array · optionnel
    Alternative derivation rules (rule set B)
  • blueprint · string · optionnel
    Blueprint name (workflow_name) to use
  • constraints_b · array · optionnel
    Alternative constraints (rule set B)
  • structured_data · object · requis
    The document's extracted fields as key/value pairs. Keys are open by design - your Blueprint's rules define what is checked

  Exemple d'appel

  {"name": "counterfactual", "arguments": {"api_key": "texte", "structured_data": {}}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• create_blueprint≈ 917 tokens

  Create a Blueprint - the governance contract validation runs against. A Blueprint defines what correct means for your data: fields, the math that must hold between them, and acceptable ranges. Start from load_rule_pack or discover_patterns if you have no rules yet; invoke the blueprint_guide prompt for the full rule/constraint reference. Returns the new Blueprint's API key.

  Paramètres, retour & exemple (16)

  Paramètres

  • mode · string · optionnel · observe | enforce
    observe: platform checks the agent's work; enforce: platform computes derived fields itself
  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • require_math · boolean · optionnel
    Validate mathematical relationships
  • customer_name · string · requis
    Organization or project name (also used for storage folder naming)
  • workflow_name · string · requis
    Unique Blueprint identifier; the value passed as 'blueprint' in validate
  • derived_fields · array · optionnel
    Field names the platform computes from other fields, e.g. ['subtotal','total']
  • semantic_checks · array · optionnel
    Domain-specific semantic check objects
  • derivation_rules · array · optionnel
    Math rules as objects. Types: add, subtract, multiply, divide, round, copy, sum (multi-operand), items_multiply, items_sum. Each needs 'type' plus its fields; see the blueprint_guide prompt
  • extracted_fields · array · optionnel
    Field names the agent extracts from source data, e.g. ['vendor','qty','unit_cost']
  • require_coherence · boolean · optionnel
    Check cross-field plausibility
  • formal_constraints · array · optionnel
    Constraint objects. Types incl. magnitude_anchor {field,min,max}, relative_anchor {field,reference_field,ratio_min,ratio_max}, max_action_threshold {field,threshold,on_violation}, required_fields {fields}, equals, range, in_set, regex_match, items_magnitude_anchor; see the blueprint_guide prompt
  • require_provenance · boolean · optionnel
    Require extraction source locations for fields
  • require_consistency · boolean · optionnel
    Check internal field consistency
  • enable_drift_tracking · boolean · optionnel
    Track pattern stability across batches
  • require_high_assurance · boolean · optionnel
    Strictest mode: every check must pass
  • enable_anomaly_detection · boolean · optionnel
    Flag records that break no rules but do not fit the reference pattern

  Exemple d'appel

  {"name": "create_blueprint", "arguments": {"api_key": "texte", "customer_name": "texte", "workflow_name": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• create_chain≈ 271 tokens

  Create a multi-agent sequential chain: stages validate in order against one Blueprint, repairs propagate forward, TTL bounds the run. Siblings: submit_chain_stage advances the chain; handoff_audit verifies a transition between stages. Returns chain_id.

  Paramètres, retour & exemple (4)

  Paramètres

  • ttl · integer · optionnel
    Chain timeout in seconds; stages cannot advance after expiry
  • stages · array · requis
    Stage definitions, e.g. [{'stage_name':'extract','agent_name':'PDF Agent'}]; minimum 2
  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • blueprint · string · requis
    Blueprint governing all stages of the chain

  Exemple d'appel

  {"name": "create_chain", "arguments": {"stages": [], "api_key": "texte", "blueprint": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• decompose_failure≈ 523 tokens

  Measure the numeric difference between original and corrected values: how much lies along the supplied linear rule directions (projection) and how much lies outside them (residual), with per-field contributions. Diagnostic only: it does not establish a cause, and neither record is verified. Use with a known-correct version to diff against; use analyze_anomaly when you only have the suspicious payload. Diagnostics-tier tool.

  Paramètres, retour & exemple (6)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • blueprint · string · optionnel
    Load rules from this Blueprint instead of passing them inline
  • original_values · object · requis
    Original numeric field values as {field: number}
  • corrected_values · object · requis
    Corrected/expected numeric field values as {field: number}
  • derivation_rules · array · optionnel
    Math rules as objects. Types: add, subtract, multiply, divide, round, copy, sum (multi-operand), items_multiply, items_sum. Each needs 'type' plus its fields; see the blueprint_guide prompt
  • formal_constraints · array · optionnel
    Constraint objects. Types incl. magnitude_anchor {field,min,max}, relative_anchor {field,reference_field,ratio_min,ratio_max}, max_action_threshold {field,threshold,on_violation}, required_fields {fields}, equals, range, in_set, regex_match, items_magnitude_anchor; see the blueprint_guide prompt

  Exemple d'appel

  {"name": "decompose_failure", "arguments": {"api_key": "texte", "original_values": {}, "corrected_values": {}}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• delete_api_key≈ 271 tokens

  Permanently delete one of the caller's API keys. DESTRUCTIVE — agents using the deleted key will receive auth errors immediately. The Blueprint a key was tied to (if any) is NOT affected; only the credential is revoked. To delete a Blueprint and all its keys, use delete_blueprint. The target key can be specified two ways: - As the full key string (gai_...). - As a key_id (SHA-256 hash from list_api_keys).

  Paramètres, retour & exemple (3)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • confirm · boolean · optionnel
    Must be true to confirm this irreversible action
  • key_to_delete · string · requis
    The gai_ key to delete

  Exemple d'appel

  {"name": "delete_api_key", "arguments": {"api_key": "texte", "key_to_delete": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• delete_blueprint≈ 198 tokens

  Permanently delete a Blueprint and revoke its API keys. Irreversible; requires confirm=true. Account-level keys are unaffected.

  Paramètres, retour & exemple (3)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • confirm · boolean · optionnel
    Must be true to confirm this irreversible action
  • workflow_name · string · requis
    Blueprint to delete; its API keys are revoked

  Exemple d'appel

  {"name": "delete_blueprint", "arguments": {"api_key": "texte", "workflow_name": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• discover_patterns≈ 217 tokens

  Learn candidate validation rules and structural document types from a batch of your records, deterministically - no Blueprint required. Promote results with approve_rule. Source data is not stored.

  Paramètres, retour & exemple (3)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • blueprint · string · optionnel
    Blueprint name (workflow_name) to use
  • documents · array · requis
    List of structured records (objects) to analyze

  Exemple d'appel

  {"name": "discover_patterns", "arguments": {"api_key": "texte", "documents": []}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• forecast≈ 326 tokens

  Deterministic forward reasoning: from the current data state, generate and rank the valid next states reachable under the Blueprint's rules.

  Paramètres, retour & exemple (6)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • rank_by · string · optionnel · drift | confidence | shortest | risk
    Ranking criterion for returned paths
  • blueprint · string · optionnel
    Blueprint name (workflow_name) to use
  • max_depth · integer · optionnel
    Search depth, 1-10
  • max_branches · integer · optionnel
    Branches per step, 1-10
  • structured_data · object · requis
    The document's extracted fields as key/value pairs. Keys are open by design - your Blueprint's rules define what is checked

  Exemple d'appel

  {"name": "forecast", "arguments": {"api_key": "texte", "structured_data": {}}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• geometric_confidence≈ 196 tokens

  Summarize an already-computed state_vector into a confidence level (high/medium/low) with a recommendation. Post-hoc digest - use analyze_anomaly or check_drift for fresh analysis of raw data.

  Paramètres, retour & exemple (2)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • state_vector · object · requis
    state_vector object from a prior validate or get_execution_trace result

  Exemple d'appel

  {"name": "geometric_confidence", "arguments": {"api_key": "texte", "state_vector": {}}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• get_execution_trace≈ 252 tokens

  Run validation and return the per-node execution trace (node names, deterministic flags, timing) plus the verdict and determinism hash. Use validate for normal operation; this is for debugging and audit preparation.

  Paramètres, retour & exemple (3)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • blueprint · string · optionnel
    Blueprint name (workflow_name) to use
  • structured_data · object · requis
    The document's extracted fields as key/value pairs. Keys are open by design - your Blueprint's rules define what is checked

  Exemple d'appel

  {"name": "get_execution_trace", "arguments": {"api_key": "texte", "structured_data": {}}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• get_inference_trace≈ 240 tokens

  Retrieve the durable audit trail for a governed generation: every recorded decision and its reasons.

  Paramètres, retour & exemple (4)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • blueprint · string · optionnel
    Blueprint namespace used when recording the trace
  • inference_id · string · requis
    Caller-chosen ID grouping the steps of one generation
  • blueprint_version · string · optionnel
    Optional historical blueprint_version hash returned by govern_inference

  Exemple d'appel

  {"name": "get_inference_trace", "arguments": {"api_key": "texte", "inference_id": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• govern_inference≈ 483 tokens

  Quality-govern an in-progress AI generation step BEFORE its output is used (complements validate, which checks finished documents). Returns an action - STOP, CONTINUE, REPAIR_REGION, REUSE_MOTIF, REVIEW, ESCALATE - with a plain-language explanation. Structural scores do not establish task correctness. Check safe_to_finalize and acceptance coverage. Persistence success is reported; read traces in the same Blueprint namespace.

  Paramètres, retour & exemple (8)

  Paramètres

  • source · string · optionnel
    Free-form caller label recorded for audit
  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • payload · object · requis
    Task-type payload: generative_text {text,...}; retrieval {query,candidates}; generic {features}
  • blueprint · string · optionnel
    Owned Blueprint namespace for the trace
  • task_type · string · requis · generative_text | retrieval | generic
    Kind of generation step being governed
  • step_index · integer · optionnel
    Step number within this generation (0-based)
  • constraints · object · optionnel
    Optional governance constraint config object
  • inference_id · string · requis
    Caller-chosen ID grouping the steps of one generation

  Exemple d'appel

  {"name": "govern_inference", "arguments": {"api_key": "texte", "payload": {}, "task_type": "generative_text", "inference_id": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• handoff_audit≈ 300 tokens

  Audit a handoff between two chain stages: a context capsule of verified facts from the prior stage, and (if proposed_data is given) a compatibility verdict that catches fields mutated in transit. Siblings: create_chain, submit_chain_stage.

  Paramètres, retour & exemple (5)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • chain_id · string · requis
    Chain identifier returned by create_chain
  • to_stage · string · requis
    Stage about to start (agent B)
  • from_stage · string · requis
    Completed stage name (agent A)
  • proposed_data · object · optionnel
    Data agent B intends to submit; checked for mutation against agent A's verified fields

  Exemple d'appel

  {"name": "handoff_audit", "arguments": {"api_key": "texte", "chain_id": "texte", "to_stage": "texte", "from_stage": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• list_api_keys≈ 110 tokens

  List this account's API keys (masked) with their Blueprint bindings.

  Paramètres, retour & exemple (1)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)

  Exemple d'appel

  {"name": "list_api_keys", "arguments": {"api_key": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• list_blueprints≈ 130 tokens

  List the Blueprints on this account with field/rule/constraint counts and mode. Use the returned workflow_name as 'blueprint' in validate.

  Paramètres, retour & exemple (1)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)

  Exemple d'appel

  {"name": "list_blueprints", "arguments": {"api_key": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• load_rule_pack≈ 169 tokens

  Load a prebuilt Blueprint template (invoices, timecards, legal, POs, claims). Call without pack_id to list packs; then create_blueprint to save a customized copy.

  Paramètres, retour & exemple (2)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • pack_id · string · optionnel
    Rule pack ID; omit to list available packs

  Exemple d'appel

  {"name": "load_rule_pack", "arguments": {"api_key": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• profile_blueprint_robustness≈ 226 tokens

  Sweep the Blueprint's numeric constraint bounds and report verdict stability: the stable band, the scales where the verdict first flips, and advice. Use before deploying bound changes.

  Paramètres, retour & exemple (3)

  Paramètres

  • config · object · optionnel
    Raw blueprint config to profile (used when 'blueprint' is not given)
  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • blueprint · string · optionnel
    Blueprint name (workflow_name) to use

  Exemple d'appel

  {"name": "profile_blueprint_robustness", "arguments": {"api_key": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• recent_inference_decisions≈ 258 tokens

  Recent generation-governance decisions in this owner's Blueprint version.

  Paramètres, retour & exemple (5)

  Paramètres

  • limit · integer · optionnel
    Maximum rows to return
  • action · string · optionnel
    Optional action filter (STOP, CONTINUE, REVIEW, ...)
  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • blueprint · string · optionnel
    Blueprint namespace used when recording the trace
  • blueprint_version · string · optionnel
    Optional historical blueprint_version hash returned by govern_inference

  Exemple d'appel

  {"name": "recent_inference_decisions", "arguments": {"api_key": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• reject_rule≈ 189 tokens

  Reject a discovered candidate rule so it will not be promoted into a Blueprint. Pair with approve_rule after discover_patterns.

  Paramètres, retour & exemple (3)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • rule_id · string · requis
    Discovered rule ID from discover_patterns
  • blueprint · string · optionnel
    Blueprint name (workflow_name) to use

  Exemple d'appel

  {"name": "reject_rule", "arguments": {"api_key": "texte", "rule_id": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• repair≈ 402 tokens

  One-shot repair: return corrected values that would make failing data valid under the Blueprint. Use repair_path to see the steps instead.

  Paramètres, retour & exemple (5)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • blueprint · string · optionnel
    Blueprint name (workflow_name) to use
  • structured_data · object · requis
    The document's extracted fields as key/value pairs. Keys are open by design - your Blueprint's rules define what is checked
  • derivation_rules · array · optionnel
    Math rules as objects. Types: add, subtract, multiply, divide, round, copy, sum (multi-operand), items_multiply, items_sum. Each needs 'type' plus its fields; see the blueprint_guide prompt
  • formal_constraints · array · optionnel
    Constraint objects. Types incl. magnitude_anchor {field,min,max}, relative_anchor {field,reference_field,ratio_min,ratio_max}, max_action_threshold {field,threshold,on_violation}, required_fields {fields}, equals, range, in_set, regex_match, items_magnitude_anchor; see the blueprint_guide prompt

  Exemple d'appel

  {"name": "repair", "arguments": {"api_key": "texte", "structured_data": {}}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• repair_path≈ 328 tokens

  Find the shortest sequence of field changes taking invalid data to a valid state, as an ordered path of intermediate states. Different from repair (one-shot nearest fix): use repair_path to explain or audit the fix, or compare alternative repairs.

  Paramètres, retour & exemple (5)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • rank_by · string · optionnel · shortest | drift | confidence | risk
    Ranking criterion for returned paths
  • blueprint · string · optionnel
    Blueprint name (workflow_name) to use
  • max_depth · integer · optionnel
    Search depth, 1-10
  • structured_data · object · requis
    The document's extracted fields as key/value pairs. Keys are open by design - your Blueprint's rules define what is checked

  Exemple d'appel

  {"name": "repair_path", "arguments": {"api_key": "texte", "structured_data": {}}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• rotate_api_key≈ 160 tokens

  Replace an API key with a fresh one. The old key stops working immediately; the new key inherits its bindings.

  Paramètres, retour & exemple (2)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • key_to_rotate · string · requis
    The gai_ key to rotate; it stops working immediately

  Exemple d'appel

  {"name": "rotate_api_key", "arguments": {"api_key": "texte", "key_to_rotate": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• structural_types≈ 180 tokens

  Retrieve the document categories a discover_patterns session identified (counts, distinguishing fields, domain hints). Read-only; returns status=no_session if discovery has not run for this namespace.

  Paramètres, retour & exemple (2)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • blueprint · string · optionnel
    Blueprint name (workflow_name) to use

  Exemple d'appel

  {"name": "structural_types", "arguments": {"api_key": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• submit_chain_stage≈ 269 tokens

  Submit data for the chain's current stage; the platform validates it and advances the chain if it passes. Response includes next-stage info and accumulated repairs.

  Paramètres, retour & exemple (4)

  Paramètres

  • stage · string · requis
    Stage name to submit for (must be the chain's current stage)
  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • chain_id · string · requis
    Chain identifier returned by create_chain
  • structured_data · object · requis
    The document's extracted fields as key/value pairs. Keys are open by design - your Blueprint's rules define what is checked

  Exemple d'appel

  {"name": "submit_chain_stage", "arguments": {"stage": "texte", "api_key": "texte", "chain_id": "texte", "structured_data": {}}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• update_blueprint≈ 860 tokens

  Update an existing Blueprint in place. Only passed fields change; pass [] to clear a list. workflow_name cannot be renamed and existing API keys keep working. Different from create_blueprint: modifies an existing Blueprint, mints no new key.

  Paramètres, retour & exemple (16)

  Paramètres

  • mode · string · optionnel
    New mode: observe or enforce; omit to keep current
  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • require_math · boolean · optionnel
    Validate mathematical relationships
  • customer_name · string · optionnel
    Organization or project name (also used for storage folder naming)
  • workflow_name · string · requis
    Unique Blueprint identifier; the value passed as 'blueprint' in validate
  • derived_fields · array · optionnel
    Field names the platform computes from other fields, e.g. ['subtotal','total']
  • semantic_checks · array · optionnel
    Domain-specific semantic check objects
  • derivation_rules · array · optionnel
    Math rules as objects. Types: add, subtract, multiply, divide, round, copy, sum (multi-operand), items_multiply, items_sum. Each needs 'type' plus its fields; see the blueprint_guide prompt
  • extracted_fields · array · optionnel
    Field names the agent extracts from source data, e.g. ['vendor','qty','unit_cost']
  • require_coherence · boolean · optionnel
    Check cross-field plausibility
  • formal_constraints · array · optionnel
    Constraint objects. Types incl. magnitude_anchor {field,min,max}, relative_anchor {field,reference_field,ratio_min,ratio_max}, max_action_threshold {field,threshold,on_violation}, required_fields {fields}, equals, range, in_set, regex_match, items_magnitude_anchor; see the blueprint_guide prompt
  • require_provenance · boolean · optionnel
    Require extraction source locations for fields
  • require_consistency · boolean · optionnel
    Check internal field consistency
  • enable_drift_tracking · boolean · optionnel
    Track pattern stability across batches
  • require_high_assurance · boolean · optionnel
    Strictest mode: every check must pass
  • enable_anomaly_detection · boolean · optionnel
    Flag records that break no rules but do not fit the reference pattern

  Exemple d'appel

  {"name": "update_blueprint", "arguments": {"api_key": "texte", "workflow_name": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• validate≈ 340 tokens

  Validate structured data against a Blueprint's rules BEFORE the result is used. Returns PASS, FAIL, or REVIEW with plain-language findings, repair suggestions, a determinism hash, and a re-verifiable certificate. Same input + same rules = same verdict, every time.

  Paramètres, retour & exemple (4)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • blueprint · string · optionnel
    Blueprint name (workflow_name) to use
  • source_text · ? · optionnel
    Optional. The text of the document the fields were extracted from. When the Blueprint has a source check, extracted values are compared against this text. It is used for the check only and is not stored
  • structured_data · object · requis
    The document's extracted fields as key/value pairs. Keys are open by design - your Blueprint's rules define what is checked

  Exemple d'appel

  {"name": "validate", "arguments": {"api_key": "texte", "structured_data": {}}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• validate_repair≈ 255 tokens

  Validate structured data against a Blueprint and, when it fails, include repair suggestions (corrected values with the rule each fix is based on) in the same call. Same verdicts as validate: PASS, FAIL, or REVIEW, with reasons and proof.

  Paramètres, retour & exemple (3)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • blueprint · string · optionnel
    Blueprint name (workflow_name) to use
  • structured_data · object · requis
    The document's extracted fields as key/value pairs. Keys are open by design - your Blueprint's rules define what is checked

  Exemple d'appel

  {"name": "validate_repair", "arguments": {"api_key": "texte", "structured_data": {}}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• verify_certificate≈ 330 tokens

  Independently re-verify a validation certificate. Integrity mode checks the hash chain; full mode (certificate + original data) recomputes every attested rule from scratch - trust nothing, recheck everything.

  Paramètres, retour & exemple (5)

  Paramètres

  • data · object · optionnel
    Original payload for full re-verification; omit for integrity-only mode
  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • certificate · object · requis
    The certificate object from a prior validation response
  • derivation_rules · array · optionnel
    Rule list for independent attestation recomputation (optional)
  • formal_constraints · array · optionnel
    Optional constraints to match against the committed bundle

  Exemple d'appel

  {"name": "verify_certificate", "arguments": {"api_key": "texte", "certificate": {}}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• verify_replay≈ 255 tokens

  Check replay commitment integrity and compare recorded execution components. Version 4 includes reference context and the final result/status. A match compares commitments; this tool does not rerun the workflow or reconstruct historical reference populations, and does not prove factual correctness.

  Paramètres, retour & exemple (3)

  Paramètres

  • api_key · string · requis
    GeodesicAI API key (gai_...)
  • contract_a · object · requis
    replay_contract object from one execution
  • contract_b · object · requis
    replay_contract object to compare against contract_a

  Exemple d'appel

  {"name": "verify_replay", "arguments": {"api_key": "texte", "contract_a": {}, "contract_b": {}}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

C'est votre serveur ?

Revendiquez-le sur l'espace créateur (preuve GitHub, domaine ou npm) pour afficher le badge « propriétaire vérifié » et lancer un pentest comportemental à la demande afin d'obtenir le badge « Vérifié ».

Créer un compte / revendiquer →

Source

https://app.geodesiclabs.ai/mcp