Add Claude Code skills Memanto bridge example#581
Closed
aindrewkwk wants to merge 2 commits into
Closed
Conversation
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (5)
✅ Files skipped from review due to trivial changes (1)
🚧 Files skipped from review as they are similar to previous changes (3)
📝 WalkthroughWalkthroughThis PR adds an example project that bridges Claude Code skills and Memanto: a SkillMemory model, pluggable backends (local JSONL and Memanto SDK), API functions to recall and persist memories (pre/post hooks), a CLI, example session scripts, configuration files, and tests. ChangesClaude Code Skills Memanto Memory Bridge
Sequence DiagramsequenceDiagram
participant Skill
participant pre_skill_context
participant Backend
participant post_skill_capture
participant distill_engineering_memory
Skill->>pre_skill_context: (skill, prompt, path_hint)
pre_skill_context->>Backend: recall(query, limit=3)
Backend-->>pre_skill_context: list[SkillMemory]
pre_skill_context-->>Skill: formatted_context
Skill->>post_skill_capture: (skill, prompt, transcript, path_hint)
post_skill_capture->>distill_engineering_memory: (skill, prompt, transcript, path_hint)
distill_engineering_memory-->>post_skill_capture: SkillMemory
post_skill_capture->>Backend: remember(memory)
Backend-->>post_skill_capture: persisted
post_skill_capture-->>Skill: SkillMemory
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Poem
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
Author
|
@coderabbitai review |
✅ Actions performedReview triggered.
|
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 3
🧹 Nitpick comments (6)
examples/claudecode-skills-memanto/run_session_b.py (2)
6-6: Consider adding a docstring.A brief docstring would clarify what this example demonstrates and how it relates to Session A.
📝 Suggested docstring
def main() -> None: + """Demonstrate pre-skill context retrieval (Session B). + + Retrieves the Stripe webhook architecture decision stored by + Session A when working on a related /tdd task in the same path. + """ backend = build_backend()🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@examples/claudecode-skills-memanto/run_session_b.py` at line 6, Add a one- or two-line docstring to the top of the main() function explaining what this example demonstrates (e.g., what the script does, how it relates to Session A, and any expected inputs/outputs or behaviors) so readers quickly understand the purpose of run_session_b.py; locate the top of the main() function and insert the docstring immediately after def main() -> None: (referencing the main function) following standard Python triple-quoted string conventions.
6-15: ⚡ Quick winConsider adding error handling for robustness.
Like Session A, this script would benefit from try-except blocks around
build_backend()andpre_skill_context()to handle potential failures gracefully and demonstrate best practices.🛡️ Suggested error handling pattern
def main() -> None: - backend = build_backend() - context = pre_skill_context( - backend=backend, - skill="/tdd", - prompt="Add tests for Stripe webhook retries.", - path_hint="services/payments", - ) - print("Session B injected context:") - print(context or "No relevant engineering memory found.") + try: + backend = build_backend() + context = pre_skill_context( + backend=backend, + skill="/tdd", + prompt="Add tests for Stripe webhook retries.", + path_hint="services/payments", + ) + print("Session B injected context:") + print(context or "No relevant engineering memory found.") + except Exception as e: + print(f"Error retrieving context: {e}") + raise🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@examples/claudecode-skills-memanto/run_session_b.py` around lines 6 - 15, Wrap the backend creation and context injection in try/except blocks inside main to handle failures from build_backend() and pre_skill_context(); specifically, catch exceptions raised by build_backend() and log/print a clear error message including exception details and exit or return early, then similarly wrap pre_skill_context() to report errors and avoid printing an invalid context; ensure you reference the existing main, build_backend, and pre_skill_context symbols and keep behavior consistent with Session A (graceful failure, informative error output).examples/claudecode-skills-memanto/run_session_a.py (2)
6-6: Consider adding a docstring.Adding a brief docstring would help users understand what this example demonstrates.
📝 Suggested docstring
def main() -> None: + """Demonstrate post-skill memory capture (Session A). + + Stores a Stripe webhook architecture decision that will be + retrieved by Session B when working on related tasks. + """ backend = build_backend()🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@examples/claudecode-skills-memanto/run_session_a.py` at line 6, Add a concise docstring to the main function to explain what this example demonstrates and how to run it; specifically, update the def main() function by inserting a short triple-quoted docstring immediately under the signature (e.g., one- to three-line description of the example's purpose, inputs/outputs if any, and usage) so readers understand the example without inspecting the implementation.
6-20: ⚡ Quick winConsider adding error handling for robustness.
The script lacks try-except blocks around
build_backend()andpost_skill_capture(). For example code, wrapping these calls in error handling would make the script more robust and educational by demonstrating proper error management patterns.🛡️ Suggested error handling pattern
def main() -> None: - backend = build_backend() - memory = post_skill_capture( - backend=backend, - skill="/grill-with-docs", - prompt="Review the payments service architecture.", - transcript=( - "Decision: use a repository layer for Stripe webhook persistence. " - "Keep webhook signature verification at the HTTP boundary and store " - "only normalized event records under services/payments." - ), - path_hint="services/payments", - ) - print("Session A stored memory:") - print(f"- {memory.title}: {memory.content}") + try: + backend = build_backend() + memory = post_skill_capture( + backend=backend, + skill="/grill-with-docs", + prompt="Review the payments service architecture.", + transcript=( + "Decision: use a repository layer for Stripe webhook persistence. " + "Keep webhook signature verification at the HTTP boundary and store " + "only normalized event records under services/payments." + ), + path_hint="services/payments", + ) + print("Session A stored memory:") + print(f"- {memory.title}: {memory.content}") + except Exception as e: + print(f"Error capturing memory: {e}") + raise🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@examples/claudecode-skills-memanto/run_session_a.py` around lines 6 - 20, The script calls build_backend() and post_skill_capture() in main() without error handling; wrap those calls in a try/except block inside main (or separate try/except around each call) to catch exceptions, log or print a concise error message, and exit or re-raise as appropriate; locate the calls to build_backend and post_skill_capture in the main function and ensure exceptions include context (e.g., "failed to build backend" or "failed to store memory") so failures are handled cleanly for this example script.examples/claudecode-skills-memanto/README.md (2)
16-27: 💤 Low valueConsider adding
.gitignoreto the file listing.The directory structure omits
.gitignore, which is part of this PR and helps reviewers understand the complete project layout.📝 Suggested addition
examples/claudecode-skills-memanto/ ├── README.md ├── requirements.txt ├── .env.example +├── .gitignore ├── memanto_skills.py🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@examples/claudecode-skills-memanto/README.md` around lines 16 - 27, Update the examples/claudecode-skills-memanto/ README.md tree to include the .gitignore entry so the project layout matches the PR; edit the code block in README.md (the directory listing) to add a line for ".gitignore" (e.g., between .env.example and memanto_skills.py) and ensure the actual .gitignore file is present in the repo so the README and repository contents stay in sync.
83-89: 💤 Low valueConsider documenting ruff validation mentioned in PR description.
The PR description lists "ruff check/format" as validation commands, but the README only mentions pytest and py_compile. Including ruff commands would provide complete validation coverage.
📝 Suggested addition
```bash cd examples/claudecode-skills-memanto python -m pytest tests -q python -m py_compile memanto_skills.py run_session_a.py run_session_b.py +ruff check . +ruff format --check .</details> <details> <summary>🤖 Prompt for AI Agents</summary>Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.In
@examples/claudecode-skills-memanto/README.mdaround lines 83 - 89,
Validation section in README.md omits the ruff lint/format checks referenced in
the PR description; update the Validation block in
examples/claudecode-skills-memanto/README.md to include the ruff commands (e.g.,
add "ruff check ." and "ruff format --check .") alongside the existing pytest
and python -m py_compile commands that target memanto_skills.py,
run_session_a.py and run_session_b.py so the README reflects full validation
steps.</details> </blockquote></details> </blockquote></details> <details> <summary>🤖 Prompt for all review comments with AI agents</summary>Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.Inline comments:
In@examples/claudecode-skills-memanto/memanto_skills.py:
- Around line 55-57: The current recall logic treats any single shared token as
a match (score = len(query_terms & _tokens(haystack))) which allows common words
to trigger unrelated memories; update the matching in both places where this
appears (the score calculation and append to scored involving query_terms,
_tokens(haystack), score, and memory) to require a stricter threshold (e.g.,
require score >= 2 or score >= max(1, int(len(query_terms) * 0.3))) before
appending to scored so that at least multiple or a meaningful proportion of
query tokens must overlap to qualify as a memory match.- Around line 65-68: The JSONL parsing in the loop that builds memories (using
self.path.read_text, json.loads and SkillMemory(**...)) should be guarded so a
single corrupted/truncated line doesn't raise and abort recall; wrap the
per-line parsing and dataclass construction in a try/except (catch
json.JSONDecodeError and ValueError/TypeError from SkillMemory construction),
skip the bad line on exception (optionally log a warning with the offending
line), and continue appending only successfully parsed SkillMemory instances to
memories.- Around line 170-172: Explicit decision extraction currently concatenates
prompt and transcript (text = f"{prompt}\n{transcript}") before calling
_find_explicit_decision, which can surface prompt phrasing as a decision; change
the call to run _find_explicit_decision against the transcript only (e.g., pass
transcript or set text = transcript) so that only session output is analyzed;
update references to text/explicit accordingly (look for variable text, the
function _find_explicit_decision, and the prompt/transcript variables) and
ensure no prompt content is included in the explicit decision check.
Nitpick comments:
In@examples/claudecode-skills-memanto/README.md:
- Around line 16-27: Update the examples/claudecode-skills-memanto/ README.md
tree to include the .gitignore entry so the project layout matches the PR; edit
the code block in README.md (the directory listing) to add a line for
".gitignore" (e.g., between .env.example and memanto_skills.py) and ensure the
actual .gitignore file is present in the repo so the README and repository
contents stay in sync.- Around line 83-89: Validation section in README.md omits the ruff lint/format
checks referenced in the PR description; update the Validation block in
examples/claudecode-skills-memanto/README.md to include the ruff commands (e.g.,
add "ruff check ." and "ruff format --check .") alongside the existing pytest
and python -m py_compile commands that target memanto_skills.py,
run_session_a.py and run_session_b.py so the README reflects full validation
steps.In
@examples/claudecode-skills-memanto/run_session_a.py:
- Line 6: Add a concise docstring to the main function to explain what this
example demonstrates and how to run it; specifically, update the def main()
function by inserting a short triple-quoted docstring immediately under the
signature (e.g., one- to three-line description of the example's purpose,
inputs/outputs if any, and usage) so readers understand the example without
inspecting the implementation.- Around line 6-20: The script calls build_backend() and post_skill_capture() in
main() without error handling; wrap those calls in a try/except block inside
main (or separate try/except around each call) to catch exceptions, log or print
a concise error message, and exit or re-raise as appropriate; locate the calls
to build_backend and post_skill_capture in the main function and ensure
exceptions include context (e.g., "failed to build backend" or "failed to store
memory") so failures are handled cleanly for this example script.In
@examples/claudecode-skills-memanto/run_session_b.py:
- Line 6: Add a one- or two-line docstring to the top of the main() function
explaining what this example demonstrates (e.g., what the script does, how it
relates to Session A, and any expected inputs/outputs or behaviors) so readers
quickly understand the purpose of run_session_b.py; locate the top of the main()
function and insert the docstring immediately after def main() -> None:
(referencing the main function) following standard Python triple-quoted string
conventions.- Around line 6-15: Wrap the backend creation and context injection in
try/except blocks inside main to handle failures from build_backend() and
pre_skill_context(); specifically, catch exceptions raised by build_backend()
and log/print a clear error message including exception details and exit or
return early, then similarly wrap pre_skill_context() to report errors and avoid
printing an invalid context; ensure you reference the existing main,
build_backend, and pre_skill_context symbols and keep behavior consistent with
Session A (graceful failure, informative error output).</details> <details> <summary>🪄 Autofix (Beta)</summary> Fix all unresolved CodeRabbit comments on this PR: - [ ] <!-- {"checkboxId": "4b0d0e0a-96d7-4f10-b296-3a18ea78f0b9"} --> Push a commit to this branch (recommended) - [ ] <!-- {"checkboxId": "ff5b1114-7d8c-49e6-8ac1-43f82af23a33"} --> Create a new PR with the fixes </details> --- <details> <summary>ℹ️ Review info</summary> <details> <summary>⚙️ Run configuration</summary> **Configuration used**: defaults **Review profile**: CHILL **Plan**: Pro Plus **Run ID**: `ec059a79-e3ec-4ddd-b633-0e94ac48398b` </details> <details> <summary>📥 Commits</summary> Reviewing files that changed from the base of the PR and between 01b8b24151e7efa9f56fe9979fa8e8dd54c9c455 and a4a606f8c557d20903e8a8a9cbfe58386822e9b3. </details> <details> <summary>📒 Files selected for processing (9)</summary> * `examples/claudecode-skills-memanto/.env.example` * `examples/claudecode-skills-memanto/.gitignore` * `examples/claudecode-skills-memanto/README.md` * `examples/claudecode-skills-memanto/memanto_skills.py` * `examples/claudecode-skills-memanto/requirements.txt` * `examples/claudecode-skills-memanto/run_session_a.py` * `examples/claudecode-skills-memanto/run_session_b.py` * `examples/claudecode-skills-memanto/tests/conftest.py` * `examples/claudecode-skills-memanto/tests/test_memanto_skills.py` </details> </details> <!-- This is an auto-generated comment by CodeRabbit for review status -->
Collaborator
|
We were blown away by the community's creativity and the sheer volume of high-quality submissions! After reviewing all the pull requests against the bounty's success matrix, we have decided to move forward with merging PR #692, which implemented a highly portable prompt-injection architecture via CLAUDE.md. We are closing this PR because it falls into one of the architectural approaches that we ultimately decided against for the ecosystem:
We deeply appreciate the time and engineering effort you put into this submission. The codebase was fantastic to review, and we hope to see you in future Moorcheh bounties! |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
/claim #508
Bounty: https://www.bountyhub.dev/bounty/view/3a63800c-7c18-41d2-870f-c62344f8a3fe
What it adds:
examples/claudecode-skills-memanto, a Claude Code Skills-style memory bridge for Memanto.prehook: recalls task/path-relevant engineering memory before a skill starts.posthook: distills completed skill output into durable project memory.MEMANTO_SKILLS_BACKEND=memanto./tddsession receiving it as injected context.Social showcase:
Validation run locally:
uv run python -m py_compile examples/claudecode-skills-memanto/memanto_skills.py examples/claudecode-skills-memanto/run_session_a.py examples/claudecode-skills-memanto/run_session_b.pyuv run pytest examples/claudecode-skills-memanto/tests -quv run ruff check examples/claudecode-skills-memantouv run ruff format --check examples/claudecode-skills-memantouv run python examples/claudecode-skills-memanto/run_session_a.py && uv run python examples/claudecode-skills-memanto/run_session_b.pySummary by CodeRabbit
New Features
Documentation
Tests
Chores