feat: add rules management with 5 new MCP tools#75
Conversation
📝 WalkthroughSummary by CodeRabbit
WalkthroughAdds Firefly III rule-management: new FireflyClient rule endpoints, Pydantic request/response models, a RuleService with search/update/test/execute flows, an MCP "rules" server wiring, and corresponding unit and integration tests. Changes
Sequence Diagram(s)sequenceDiagram
participant Client as MCP Client
participant Tools as Rules Server
participant Service as RuleService
participant Firefly as FireflyClient
Client->>Tools: call test_rule(rule_id, start, end, accounts)
Tools->>Service: test_rule(TestRuleRequest)
Service->>Firefly: get_rule(rule_id)
Firefly-->>Service: RuleSingle (metadata)
Service->>Firefly: test_rule(rule_id, start, end, accounts)
Firefly-->>Service: TransactionArray (matches)
Service->>Service: convert transactions -> Transaction models
Service-->>Tools: RuleTestResult (count, matches)
Tools-->>Client: return RuleTestResult
sequenceDiagram
participant Client as MCP Client
participant Tools as Rules Server
participant Service as RuleService
participant Firefly as FireflyClient
Client->>Tools: call execute_rule(rule_id, start, end, confirm=true)
Tools->>Service: execute_rule(ExecuteRuleRequest)
Service->>Service: validate confirm == True
Service->>Firefly: get_rule(rule_id)
Firefly-->>Service: RuleSingle (metadata)
Service->>Firefly: trigger_rule(rule_id, start, end, accounts)
Firefly-->>Service: HTTP 204 -> boolean success
Service-->>Tools: RuleExecuteResult (success, message)
Tools-->>Client: return execution result
Estimated code review effort🎯 4 (Complex) | ⏱️ ~45 minutes Possibly related PRs
Poem
🚥 Pre-merge checks | ✅ 3✅ Passed checks (3 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 5
🤖 Fix all issues with AI agents
In `@src/lampyrid/clients/firefly.py`:
- Around line 503-532: The rule-related APIs and their request models currently
use account_ids: Optional[list[str]] (e.g., test_rule and
ExecuteRuleRequest/TestRuleRequest) while insight methods use
Optional[list[int]]; unify them by changing all rule method signatures, query
param handling, and request models to use Optional[list[int]] (or pick int
consistently) so the shared query key 'accounts[]' always receives integers;
update test_rule, trigger_rule, TestRuleRequest, ExecuteRuleRequest and any
other rule-insight callers to accept/serialize list[int] and ensure
params['accounts[]'] passes ints (or converts incoming ints to strings only at
the HTTP layer if necessary).
In `@src/lampyrid/services/rules.py`:
- Around line 243-245: The current assignments use the faulty "or True"/"or
False" pattern which coerces falsy values like False to the defaults; update the
assignments to use explicit None-coalescing so an explicit False is preserved.
For example, replace the active and stop_processing assignments that reference
rule_attrs.active and rule_attrs.stop_processing with logic that checks "is
None" (e.g., set active = True if rule_attrs.active is None else
rule_attrs.active and set stop_processing = False if rule_attrs.stop_processing
is None else rule_attrs.stop_processing), leaving strict as rule_attrs.strict;
target the active/stop_processing assignments in the function/class where these
lines appear.
- Around line 135-137: The current assignment uses boolean "or" which forces
True for active and may mis-handle stop_processing; update the mappings where
you set active, strict, and stop_processing from rule_read.attributes to
explicitly check for None and only use the default when the attribute is None
(e.g., use a conditional that returns rule_read.attributes.active if it is not
None else True, and similarly for stop_processing defaulting to False), keeping
the field names (rule_read, attributes, active, strict, stop_processing) to
locate and modify the code.
- Around line 175-177: The assignment active=rule_attrs.active or True
incorrectly forces active to True even when rule_attrs.active is False; change
it to preserve explicit False by using a None-check such as active =
rule_attrs.active if rule_attrs.active is not None else True (or equivalent like
getattr(rule_attrs, "active", True) with an explicit None check) in the code
that constructs the rule (the lines referencing rule_attrs.active,
rule_attrs.strict, rule_attrs.stop_processing) so inactive rules remain
inactive.
In `@tests/unit/test_rules_service.py`:
- Around line 148-163: The test test_search_rules_by_trigger_type uses an
invalid trigger_type value 'amount_greater_than' when creating rule2 via
_make_rule_read; update the trigger_type to a valid RuleTriggerKeyword such as
'amount_more' so the mock rule matches the service's expected enums and the
SearchRulesRequest(trigger_type='description') test logic remains correct.
🧹 Nitpick comments (2)
src/lampyrid/models/lampyrid_models.py (1)
1190-1206: Consider adding a validator for trigger/action structure.The
triggersandactionsfields useList[dict[str, Any]]which defers validation to the service layer. While this works, adding a basic validator to check for required keys (liketype) at the model level would provide earlier feedback to users.♻️ Optional: Add basic structure validation
+ `@model_validator`(mode='after') + def validate_trigger_action_structure(self): + """Validate that triggers/actions have required 'type' field.""" + if self.triggers: + for i, t in enumerate(self.triggers): + if 'type' not in t: + raise ValueError(f"triggers[{i}] missing required 'type' field") + if self.actions: + for i, a in enumerate(self.actions): + if 'type' not in a: + raise ValueError(f"actions[{i}] missing required 'type' field") + return selfsrc/lampyrid/services/rules.py (1)
228-233: Move inline import to module level.The
RuleActionUpdateimport is done inline within the function. While functional, this is inconsistent withRuleTriggerUpdatewhich is imported at the module level.♻️ Proposed refactor
At line 13, add
RuleActionUpdateto the existing import:-from ..models.firefly_models import RuleTriggerUpdate, RuleUpdate +from ..models.firefly_models import RuleActionUpdate, RuleTriggerUpdate, RuleUpdateThen remove the inline import at line 229.
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Fix all issues with AI agents
In `@src/lampyrid/models/lampyrid_models.py`:
- Around line 1223-1244: Add a Pydantic validation to both TestRuleRequest and
ExecuteRuleRequest to ensure start_date <= end_date: implement a `@root_validator`
(or classmethod validator) in each model that checks the two fields and raises a
ValueError (or ValidationError) with a clear message when start_date is after
end_date so API returns a validation error instead of empty results; locate the
models by name (TestRuleRequest, ExecuteRuleRequest) in lampyrid_models.py and
add the validator logic to each class.
🧹 Nitpick comments (2)
src/lampyrid/models/lampyrid_models.py (2)
1191-1207: Consider using typed models fortriggersandactionsinstead ofdict[str, Any].Using untyped dicts forgoes compile-time and validation-time safety. Invalid payloads (e.g. missing
typekey, wrong value types) will only surface as downstream API errors rather than clear Pydantic validation failures. A lightweight typed model (even aTypedDictor a small Pydantic model withtype: strrequired andvalue: Optional[str]) would catch malformed input earlier and provide better error messages to MCP clients.That said, the current approach does offer flexibility for passing API-specific fields that may not all be modeled yet, so this is a tradeoff.
1215-1221: Duplicated_parse_string_inputvalidator acrossTestRuleRequestandExecuteRuleRequest.Both classes define identical
mode='before'validators. If more request models withdatefields are added, this pattern will spread further. Consider extracting a shared mixin or base class:♻️ Example DRY refactor
+class _JsonStringParseMixin(BaseModel): + """Mixin for models that need to handle JSON-string input from MCP clients.""" + + `@model_validator`(mode='before') + `@classmethod` + def _parse_string_input(cls, data: Any) -> Any: + if isinstance(data, str): + return json.loads(data) + return data + + -class TestRuleRequest(BaseModel): +class TestRuleRequest(_JsonStringParseMixin): """Request model for testing a rule (preview matches).""" model_config = ConfigDict(extra='forbid') - `@model_validator`(mode='before') - `@classmethod` - def _parse_string_input(cls, data: Any) -> Any: - """Handle MCP clients that serialize the request as a JSON string.""" - if isinstance(data, str): - return json.loads(data) - return data - rule_id: str = Field(...) ... -class ExecuteRuleRequest(BaseModel): +class ExecuteRuleRequest(_JsonStringParseMixin): """Request model for executing a rule (apply changes).""" model_config = ConfigDict(extra='forbid') - `@model_validator`(mode='before') - `@classmethod` - def _parse_string_input(cls, data: Any) -> Any: - """Handle MCP clients that serialize the request as a JSON string.""" - if isinstance(data, str): - return json.loads(data) - return data - rule_id: str = Field(...) ...Also applies to: 1252-1258
There was a problem hiding this comment.
Actionable comments posted: 3
🤖 Fix all issues with AI agents
In `@src/lampyrid/models/lampyrid_models.py`:
- Around line 1215-1221: The pre-validator _parse_string_input currently calls
json.loads(data) unguarded; wrap that call in a try/except that catches
json.JSONDecodeError and re-raises a plain ValueError (or TypeError) with a
clear message so Pydantic will convert it into a ValidationError; update this
logic in both classes that define _parse_string_input (TestRuleRequest and
ExecuteRuleRequest) so malformed JSON strings produce a validation-friendly
error instead of an uncaught JSONDecodeError.
In `@src/lampyrid/services/rules.py`:
- Around line 84-107: Extract the duplicated mapping logic into a single private
helper method (e.g., _to_rule) that accepts the Firefly response object (handle
both rule_single.data and rule_read.attributes cases) and returns a Rule; inside
the helper read rule_attrs = ...attributes, map
id/title/description/trigger/actions/triggers to Rule, replacing brittle "or
True"/"or False" with explicit None checks (e.g., active = rule_attrs.active if
rule_attrs.active is not None else True; prohibited = t.prohibited if
t.prohibited is not None else False; stop_processing likewise) and construct
nested RuleTriggerSimple and RuleActionSimple; then update create_rule,
search_rules, get_rule, and update_rule to call self._to_rule(...) (for
search_rules map each result through the helper).
- Around line 67-72: Change RuleStore.rule_group_id to be Optional[str] in the
Firefly model (update its type annotation and any serializers/deserializers) and
then modify the code that builds the RuleStore instance in the rules service so
it sets rule_group_id to '1' only when rule_group_title is not provided (i.e.,
rule_group_id = '1' if not req.rule_group_title else None). Update any
imports/uses of RuleStore and ensure RuleUpdate typing remains compatible.
🧹 Nitpick comments (7)
src/lampyrid/models/lampyrid_models.py (3)
1148-1163:SearchRulesRequestvalidator treats empty strings as "no criterion" — reasonable but worth noting.The truthiness-based check (
any([self.trigger_type, ...])) means passingtrigger_type=""would not count as a criterion. This is acceptable behavior since an empty string isn't a meaningful filter, but it differs from theSearchTransactionsRequestvalidator (lines 476-503) which explicitly handles this withisinstance(field, str) and field.strip() != ''. The inconsistency is minor but could confuse future maintainers.
1191-1207:triggersandactionstyped asList[dict[str, Any]]— validation deferred to service layer.This is a deliberate loose-typing choice, but it means invalid trigger/action structures won't be caught until the service layer calls
RuleTriggerStore(**t)/RuleActionStore(**a). ForUpdateRuleRequestandCreateRuleRequest, consider usingList[RuleTriggerStore]/List[RuleActionStore](or their update variants) to shift validation left to the model boundary. This is a tradeoff with MCP tool flexibility, so flagging as optional.Also applies to: 1340-1357
1324-1331: UseRuleTriggerTypeenum instead ofstrfor thetriggerfield to improve type safety.The
triggerfield accepts only'store-journal','update-journal', or'manual-activation', but is typed asstr. Invalid values fail later at the service layer whenRuleTriggerType(req.trigger)is called, producing less informative errors. SinceRuleTriggerTypealready exists infirefly_modelsand other enum types from that module are already used in this file, constraining the field to the enum would catch invalid values at the model boundary and align with existing patterns.♻️ Proposed fix
- trigger: str = Field( - 'store-journal', + trigger: RuleTriggerType = Field( + RuleTriggerType.store_journal, description=( 'When the rule should fire. ' '"store-journal" = on new transactions (most common), ' '"update-journal" = on transaction updates.' ), )Add
RuleTriggerTypeto the imports fromfirefly_modelsat the top of the file.src/lampyrid/services/rules.py (4)
295-301: Inline import ofRuleActionUpdateis inconsistent with top-level imports.
RuleTriggerUpdateis imported at the top of the file (line 13), butRuleActionUpdateis imported inline here. Move it to the top-level import for consistency.♻️ Proposed fix
At line 13:
-from ..models.firefly_models import RuleActionStore, RuleStore, RuleTriggerStore, RuleTriggerType, RuleTriggerUpdate, RuleUpdate +from ..models.firefly_models import RuleActionStore, RuleActionUpdate, RuleStore, RuleTriggerStore, RuleTriggerType, RuleTriggerUpdate, RuleUpdateAt lines 295-301:
if req.actions is not None: try: - from ..models.firefly_models import RuleActionUpdate - rule_update.actions = [RuleActionUpdate(**a) for a in req.actions] except ValidationError as e: raise ValueError(f'Invalid action format: {e}')
141-154: Unbounded pagination loop has no safety limit.If the API returns inconsistent pagination metadata (e.g.,
current_pagenever reachingtotal_pagesdue to concurrent rule creation), this loop runs indefinitely. Consider adding a max page limit.🛡️ Proposed fix
all_rules = [] page = 1 - while True: + max_pages = 100 # Safety limit + while page <= max_pages: rule_array = await self._client.get_rules(page) all_rules.extend(rule_array.data) # Check pagination safely (can be None) if ( not rule_array.meta.pagination or rule_array.meta.pagination.current_page >= rule_array.meta.pagination.total_pages ): break page += 1
332-361:test_rulemakes two sequential API calls; the first (get_rule) is only for the title.The extra round-trip to fetch the rule title is unnecessary latency. Consider whether the title is required in the result, or if it could be populated from the test response itself. If the title is essential, this is acceptable — just noting the tradeoff.
125-139: User-supplied regex compiled without bounds — low ReDoS risk but worth noting.User-provided
trigger_value_patternandaction_value_patternare compiled directly. In an MCP context, the user is typically the LLM agent or a trusted client, so the risk is minimal. If this were ever exposed to untrusted input, consider adding a timeout or pattern complexity limit.
|
Hi @edmundhighcock, Apologies for the late reply and thank you for your contribution! Thanks! |
Add comprehensive rules functionality to manage Firefly III rules: - search_rules: Client-side filtering with regex pattern support - get_rule: Retrieve single rule details - update_rule: Modify rule configuration (triggers/actions) - test_rule: Preview rule matches without changes (read-only) - execute_rule: Apply changes with explicit confirmation requirement Implementation includes: - RuleService in services layer with client-side search/filtering - 11 new request/response models for type safety - 5 new HTTP client methods for Firefly III rule APIs - Comprehensive unit tests (19 test cases) - Safety-first design: test before execute, confirmation required This brings the total tool count from 22 to 27. Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
edmundhighcock
force-pushed
the
feat/add-rules-management
branch
from
ffd5f40 to
470f5e9
Compare
|
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #75 +/- ##
==========================================
+ Coverage 97.66% 97.75% +0.09%
==========================================
Files 19 21 +2
Lines 3121 3348 +227
==========================================
+ Hits 3048 3273 +225
- Misses 73 75 +2
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
- Fix `active or True` logic error that made inactive rules appear active - Replace all `or <default>` patterns with explicit None-coalescing - Extract Rule.from_rule_read() factory method (DRY: 3x duplication → 1) - Fix test helper using wrong enum (RuleTriggerType vs RuleTriggerKeyword) - Use Pydantic's Field(deprecated=...) for the trigger field - Move RuleActionUpdate to top-level import Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
There was a problem hiding this comment.
Actionable comments posted: 3
♻️ Duplicate comments (1)
src/lampyrid/models/lampyrid_models.py (1)
1249-1269:⚠️ Potential issue | 🟡 MinorReject inverted date ranges up front.
start_date > end_dateis still accepted for both preview and execute requests, so a caller typo turns into an empty preview or a no-op execution instead of a validation error. Add anaftervalidator on both models before these requests reach the service.Also applies to: 1278-1306
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/lampyrid/models/lampyrid_models.py` around lines 1249 - 1269, Add a post-validation check to the request models that declare start_date and end_date: implement a `@root_validator` (or an `@validator` on end_date that accesses start_date) that raises a ValueError when start_date > end_date so inverted date ranges are rejected; apply this to both the preview and execute request models that define the start_date and end_date fields (validate the start_date and end_date attributes and raise a clear error message like "start_date must be on or before end_date").
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@src/lampyrid/models/lampyrid_models.py`:
- Around line 1243-1246: The TestRuleRequest Pydantic model is being collected
by pytest due to the "Test*" class pattern; add a class attribute __test__ =
False inside the TestRuleRequest class (e.g., directly under the docstring and
above model_config) so pytest will skip collecting it, leaving the class name
unchanged and keeping model_config = ConfigDict(extra='forbid') intact.
- Around line 1224-1240: The triggers and actions fields currently use
Optional[List[dict[str, Any]]] which allows malformed payloads and silent
dropping of unknown fields; change these to use the dedicated nested request
models (e.g., replace types with Optional[List[RuleTriggerUpdate]] and
Optional[List[RuleActionUpdate]]), ensure the RuleTriggerUpdate and
RuleActionUpdate classes mark the type field as required and set Config.extra =
'forbid' (or equivalent Pydantic config) so unknown fields are rejected, and
update any validators/usage in lampyrid_models.py to reflect the new types so
invalid payloads like {} or {"prohibited": true} fail validation immediately.
In `@src/lampyrid/services/rules.py`:
- Around line 106-121: The code currently excludes empty-string trigger/action
values by using truthy checks (e.g., list comprehensions in the trigger_values
and action_values where you use "if t.value" and "if a.value"), which prevents
regexes like ^$ or .* from matching intentionally-blank values; update the
filters to only exclude None (use explicit None checks such as "if t.value is
not None" and "if a.value is not None") so empty strings remain in
trigger_values and action_values for pattern matching while still filtering out
missing values in rule_attrs.triggers and rule_attrs.actions.
---
Duplicate comments:
In `@src/lampyrid/models/lampyrid_models.py`:
- Around line 1249-1269: Add a post-validation check to the request models that
declare start_date and end_date: implement a `@root_validator` (or an `@validator`
on end_date that accesses start_date) that raises a ValueError when start_date >
end_date so inverted date ranges are rejected; apply this to both the preview
and execute request models that define the start_date and end_date fields
(validate the start_date and end_date attributes and raise a clear error message
like "start_date must be on or before end_date").
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: 3be0f6a3-291a-41ea-9167-504801ffd06e
📒 Files selected for processing (6)
src/lampyrid/clients/firefly.pysrc/lampyrid/models/lampyrid_models.pysrc/lampyrid/services/rules.pysrc/lampyrid/tools/__init__.pysrc/lampyrid/tools/rules.pytests/unit/test_rules_service.py
🚧 Files skipped from review as they are similar to previous changes (2)
- tests/unit/test_rules_service.py
- src/lampyrid/clients/firefly.py
…r rules - Fix test_search_rules_by_trigger_type: use valid enum value amount_more - Fix invalid trigger/action dict tests: use invalid enum value instead of extra fields (generated models silently ignore extra fields) - Fix empty-string filtering: use `is not None` instead of truthiness check so regex patterns can match empty trigger/action values - Add start_date <= end_date validation to TestRuleRequest and ExecuteRuleRequest with model_validator - Add 13 integration tests for rules (search, get, update, test, execute) - Add 'rules' pytest marker to pyproject.toml Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
There was a problem hiding this comment.
Actionable comments posted: 1
🧹 Nitpick comments (1)
tests/integration/test_rules.py (1)
15-48: Test helpers access private_clientmember.The helpers directly access
firefly_client._clientto make raw HTTP calls. While acceptable for integration tests that need to bypass the public API, consider adding a public method toFireflyClientfor raw requests if this pattern expands, or document that these helpers are test-only utilities.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@tests/integration/test_rules.py` around lines 15 - 48, The test helpers _create_rule_via_api and _delete_rule_via_api directly access the private attribute firefly_client._client; add a public test-only method on FireflyClient (e.g., request_raw or api_request) that wraps the internal HTTP client and exposes post/delete and raise_for_status behavior, then update _create_rule_via_api and _delete_rule_via_api to call FireflyClient.request_raw (or the chosen method) instead of accessing _client; alternatively, if you prefer not to change the class, document these helpers as test-only and justify the private access in the test module's docstring.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@src/lampyrid/services/rules.py`:
- Around line 79-84: The pagination loop can raise TypeError when
pagination.current_page or pagination.total_pages is None; update the loop
condition in the block using rule_array.meta.pagination to guard against None by
treating missing values as terminal (i.e., break if pagination is missing OR
current_page is None OR total_pages is None OR current_page >= total_pages).
Locate the check referencing rule_array.meta.pagination.current_page and
rule_array.meta.pagination.total_pages and add explicit None checks before
performing the >= comparison (ensure variable page increment remains unchanged).
---
Nitpick comments:
In `@tests/integration/test_rules.py`:
- Around line 15-48: The test helpers _create_rule_via_api and
_delete_rule_via_api directly access the private attribute
firefly_client._client; add a public test-only method on FireflyClient (e.g.,
request_raw or api_request) that wraps the internal HTTP client and exposes
post/delete and raise_for_status behavior, then update _create_rule_via_api and
_delete_rule_via_api to call FireflyClient.request_raw (or the chosen method)
instead of accessing _client; alternatively, if you prefer not to change the
class, document these helpers as test-only and justify the private access in the
test module's docstring.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: cf238a46-16fa-4ba4-a259-564e75b607c7
📒 Files selected for processing (5)
pyproject.tomlsrc/lampyrid/models/lampyrid_models.pysrc/lampyrid/services/rules.pytests/integration/test_rules.pytests/unit/test_rules_service.py
✅ Files skipped from review due to trivial changes (1)
- pyproject.toml
🚧 Files skipped from review as they are similar to previous changes (1)
- tests/unit/test_rules_service.py
The Firefly III API returns 500 when creating rules with only rule_group_title. Fix by fetching/creating a rule group first and using its ID. Also improve error reporting on creation failure. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
There was a problem hiding this comment.
Actionable comments posted: 3
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@tests/integration/test_rules.py`:
- Around line 441-472: The test currently creates a real rule via
_create_rule_via_api (rule_id) that matches transactions containing "Seed:" and
then executes it, leaving tags on persisted transactions; change the test to
avoid side effects by using a no-match trigger_value (e.g., set trigger_value to
a unique random string like "NO_MATCH_<uuid>" or "THIS_SHOULD_NOT_MATCH") when
calling _create_rule_via_api in the test (tests/integration/test_rules.py) so
the execute_rule call (mcp_client.call_tool 'execute_rule' with rule_id,
start_date, end_date, confirm) runs without mutating real transactions;
alternatively, if you prefer the other approach, create a throwaway transaction
before execute, run the rule against that transaction, then remove the
transaction and ensure rule_cleanup removes the rule—update references to
rule_id, rule_cleanup, and the execute_rule call accordingly.
- Around line 356-361: The test currently asserts that
test_result['matched_transaction_count'] is an int and then tautologically
checks it is >= 0; to ensure the preview actually matched the seeded "Seed:"
transactions change the second assertion to require a positive match
(test_result['matched_transaction_count'] > 0). Update the assertion that
references test_result['matched_transaction_count'] so the test will fail if no
transactions were matched, thereby guaranteeing the preview matched seeded data.
- Around line 28-39: The current lookup caches groups[0]['id'] which may pick an
arbitrary existing group; change the logic in the helper that sets
_test_rule_group_id to search the fetched groups list for an item whose 'title'
equals "Test Rules" and use that id if found, only issuing the POST to
firefly_client._client.post('/api/v1/rule-groups', json={'title': 'Test Rules',
'order': 1}) to create the group when no matching title exists; update any
references to _test_rule_group_id to use the found-or-created id.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: f05f04a8-451f-4627-9f1d-8867db79b178
📒 Files selected for processing (1)
tests/integration/test_rules.py
|
Hi @RadCod3, just a heads-up — integration tests have been added as requested. The latest push includes:
CI is all green. Let me know if there's anything else you'd like changed! |
|
Hey @edmundhighcock I marked the coderabbit stuff as resolved and will proceed to merge this. |
3 participants
Hi! Thanks for a great tool. I really needed rules functionality so I've added it to my fork. Hoping you might be happy to add it to the main branch.
Note: I've also added a one-click install to add this to home assistant, which many people use to host firefly iii. (See https://github.com/edmundhighcock/hassio-repository/tree/main/fireflyiii-mcp).
Claude summary
Add comprehensive rules functionality to manage Firefly III rules:
Implementation includes:
This brings the total tool count from 22 to 27.