Loading .claude/CLAUDE.md +7 −0 Changes for .claude/CLAUDE.md: 7 added lines, 0 removed lines. Original line number Diff line number Diff line Loading @@ -183,6 +183,13 @@ Scopes (optional): API group or subsystem, e.g. `subscription`, `temporal`, `rob --- ## Reviewing a Merge Request Write MR reviews in the format of `.claude/review-template.md`. Diff against the MR's target branch, not `develop`: stacked MRs target another MR's branch. --- ## What Agents Must Never Do - Hardcode broker URLs or port numbers — always use variables from `resources/variables.py`. Loading .claude/commands/review-tc.md +44 −3 Changes for .claude/commands/review-tc.md: 44 added lines, 3 removed lines. Original line number Diff line number Diff line Loading @@ -37,9 +37,10 @@ robocop format --check <file> PASS if exit code 0 (no changes needed). FAIL otherwise — list lines that would change. ### 3. Test case identifier pattern ### 3. Test case identifier pattern and file layout Every test case name inside the file must match `XXX_YY_NN <Description>` where `XXX`, `YY`, `NN` are numeric. Every test case name inside the file must match `XXX_YY_NN <Description>` where `XXX`, `YY`, `NN` are numeric, and the file must sit at `TP/NGSI-LD/<Group>/<Subgroup>/<Operation>/<TC_ID>.robot`. ### 4. Required tags on every permutation Loading Loading @@ -103,6 +104,39 @@ PYTHONPATH=./doc python -m unittest discover -s doc/tests PASS if the output ends with `OK`. FAIL with the failing tests otherwise. ### 11. Spec clause tags and asserted behaviour For every `section_reference` tag, read the clause in the spec (TS 104 175 for `core_`, TS 104 176 for `http_`, TS 104 243 for `mqtt_`): - The clause must describe the operation or behaviour the permutation tests (clause numbers don't change across spec versions, so a mismatch is a typo in the tag). - What the TC asserts (status codes, response body, notification content) must match what the clause requires. FAIL with the clause and a quote of the spec text otherwise. ### 12. Release tag Every permutation in the file uses the same `since_vX.Y.Z` or `since_ts104_vX.Y.Z` tag (an untagged permutation counts as v1.3.1), and it matches the spec version that introduced the feature. FAIL if the file mixes versions or the tag is wrong. ### 13. Robot dry run ```bash python3 -m robot --dryrun --outputdir ./results <file> ``` PASS if all TCs pass. FAIL with the errors otherwise. ### 14. Run against a broker (optional) Run the TC with `/run-suite`. Report the broker and its version, then for each failure the expected result, the result you got, its cause, and whether it is a known gap of that broker. N/A if not run. --- ## Summary Loading @@ -113,7 +147,7 @@ After all checks, output a summary table: |---|---|---| | 1 | Commit message format | PASS / FAIL / N/A | | 2 | Robocop formatting | PASS / FAIL | | 3 | TC identifier pattern | PASS / FAIL | | 3 | TC identifier pattern and file layout | PASS / FAIL | | 4 | Required tags | PASS / FAIL | | 5 | Keywords registered in doc/analysis/ | PASS / FAIL / WARN | | 6 | Setup/Teardown symmetry | PASS / WARN / N/A | Loading @@ -121,6 +155,13 @@ After all checks, output a summary table: | 8 | No hardcoded URLs | PASS / WARN | | 9 | Documentation still generates | PASS / FAIL | | 10 | Documentation unit tests | PASS / FAIL | | 11 | Spec clause tags and asserted behaviour | PASS / FAIL | | 12 | Release tag | PASS / FAIL | | 13 | Robot dry run | PASS / FAIL | | 14 | Run against a broker | PASS / WARN / N/A | If all items are PASS or WARN, state: "Ready to open a merge request." If any item is FAIL, list the remediation steps required before the MR can be opened. When reviewing someone else's MR rather than your own change, write the review in the format of `.claude/review-template.md`. .claude/review-template.md 0 → 100644 +36 −0 Changes for .claude/review-template.md: 36 added lines, 0 removed lines. Original line number Diff line number Diff line # Review — MR !<N>: <MR title> <MR URL> — reviewed at `<short sha>` against `<target branch>` (stacked on !<M>, if the target is another MR's branch). **Verdict:** <Approve | Approve after fixes (items #…) | Changes requested (items #…)> <One or two sentences: what the MR adds, and the main thing the submitter must do, if anything.> ## Checks The summary table of `/review-tc` (`.claude/commands/review-tc.md`), run on the MR's files and commits, with a short note per result (e.g. the clauses verified, `n/n OK`, the broker and its version). ## Blocking Must be fixed before merge. Omit the section if empty. ### 1. <Short title> (`path/to/file:line`) - **Problem:** <what is wrong, with spec quote or evidence> - **Fix:** <concrete change> ## Should fix Same format. Omit if empty. ## Optional / nits Same format, one or two lines each. Omit if empty. ## Broker run Omit if not run. | TC | Expected | Got | Cause | Expected failure? | |---|---|---|---|---| Loading
.claude/CLAUDE.md +7 −0 Changes for .claude/CLAUDE.md: 7 added lines, 0 removed lines. Original line number Diff line number Diff line Loading @@ -183,6 +183,13 @@ Scopes (optional): API group or subsystem, e.g. `subscription`, `temporal`, `rob --- ## Reviewing a Merge Request Write MR reviews in the format of `.claude/review-template.md`. Diff against the MR's target branch, not `develop`: stacked MRs target another MR's branch. --- ## What Agents Must Never Do - Hardcode broker URLs or port numbers — always use variables from `resources/variables.py`. Loading
.claude/commands/review-tc.md +44 −3 Changes for .claude/commands/review-tc.md: 44 added lines, 3 removed lines. Original line number Diff line number Diff line Loading @@ -37,9 +37,10 @@ robocop format --check <file> PASS if exit code 0 (no changes needed). FAIL otherwise — list lines that would change. ### 3. Test case identifier pattern ### 3. Test case identifier pattern and file layout Every test case name inside the file must match `XXX_YY_NN <Description>` where `XXX`, `YY`, `NN` are numeric. Every test case name inside the file must match `XXX_YY_NN <Description>` where `XXX`, `YY`, `NN` are numeric, and the file must sit at `TP/NGSI-LD/<Group>/<Subgroup>/<Operation>/<TC_ID>.robot`. ### 4. Required tags on every permutation Loading Loading @@ -103,6 +104,39 @@ PYTHONPATH=./doc python -m unittest discover -s doc/tests PASS if the output ends with `OK`. FAIL with the failing tests otherwise. ### 11. Spec clause tags and asserted behaviour For every `section_reference` tag, read the clause in the spec (TS 104 175 for `core_`, TS 104 176 for `http_`, TS 104 243 for `mqtt_`): - The clause must describe the operation or behaviour the permutation tests (clause numbers don't change across spec versions, so a mismatch is a typo in the tag). - What the TC asserts (status codes, response body, notification content) must match what the clause requires. FAIL with the clause and a quote of the spec text otherwise. ### 12. Release tag Every permutation in the file uses the same `since_vX.Y.Z` or `since_ts104_vX.Y.Z` tag (an untagged permutation counts as v1.3.1), and it matches the spec version that introduced the feature. FAIL if the file mixes versions or the tag is wrong. ### 13. Robot dry run ```bash python3 -m robot --dryrun --outputdir ./results <file> ``` PASS if all TCs pass. FAIL with the errors otherwise. ### 14. Run against a broker (optional) Run the TC with `/run-suite`. Report the broker and its version, then for each failure the expected result, the result you got, its cause, and whether it is a known gap of that broker. N/A if not run. --- ## Summary Loading @@ -113,7 +147,7 @@ After all checks, output a summary table: |---|---|---| | 1 | Commit message format | PASS / FAIL / N/A | | 2 | Robocop formatting | PASS / FAIL | | 3 | TC identifier pattern | PASS / FAIL | | 3 | TC identifier pattern and file layout | PASS / FAIL | | 4 | Required tags | PASS / FAIL | | 5 | Keywords registered in doc/analysis/ | PASS / FAIL / WARN | | 6 | Setup/Teardown symmetry | PASS / WARN / N/A | Loading @@ -121,6 +155,13 @@ After all checks, output a summary table: | 8 | No hardcoded URLs | PASS / WARN | | 9 | Documentation still generates | PASS / FAIL | | 10 | Documentation unit tests | PASS / FAIL | | 11 | Spec clause tags and asserted behaviour | PASS / FAIL | | 12 | Release tag | PASS / FAIL | | 13 | Robot dry run | PASS / FAIL | | 14 | Run against a broker | PASS / WARN / N/A | If all items are PASS or WARN, state: "Ready to open a merge request." If any item is FAIL, list the remediation steps required before the MR can be opened. When reviewing someone else's MR rather than your own change, write the review in the format of `.claude/review-template.md`.
.claude/review-template.md 0 → 100644 +36 −0 Changes for .claude/review-template.md: 36 added lines, 0 removed lines. Original line number Diff line number Diff line # Review — MR !<N>: <MR title> <MR URL> — reviewed at `<short sha>` against `<target branch>` (stacked on !<M>, if the target is another MR's branch). **Verdict:** <Approve | Approve after fixes (items #…) | Changes requested (items #…)> <One or two sentences: what the MR adds, and the main thing the submitter must do, if anything.> ## Checks The summary table of `/review-tc` (`.claude/commands/review-tc.md`), run on the MR's files and commits, with a short note per result (e.g. the clauses verified, `n/n OK`, the broker and its version). ## Blocking Must be fixed before merge. Omit the section if empty. ### 1. <Short title> (`path/to/file:line`) - **Problem:** <what is wrong, with spec quote or evidence> - **Fix:** <concrete change> ## Should fix Same format. Omit if empty. ## Optional / nits Same format, one or two lines each. Omit if empty. ## Broker run Omit if not run. | TC | Expected | Got | Cause | Expected failure? | |---|---|---|---|---|