Content
76%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
A dense, highly actionable operator reference with excellent token efficiency — commands and tables instead of prose. Its main weaknesses are the missing validation checkpoints in the engagement workflow (capped by the destructive-operations rule) and the JSON-RPC vs GraphQL inconsistency in the tasking section.
Suggestions
Add explicit validation checkpoints to the workflow: confirm callback and task output before lateral movement, verify persistence took hold before moving on, and validate cleanup completeness before teardown.
Reconcile the tasking model claim — pick one of 'JSON-RPC' or the GraphQL API at https://<host>:7443/v1/graphql and describe the actual interface consistently.
Verify the Python scripting example against the current mythic package API (module names and function signatures) so it is genuinely copy-paste ready.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and operator-focused: tables (agent matrix, profile matrix, framework comparison), complete CLI commands, and terse opsec bullets ('default 30s/20% is loud'). No sections explain concepts Claude already knows, and every section earns its place. It does not fall to 4 because there is no over-explanation to trim. | 5 / 5 |
Actionability | Mostly executable, copy-paste-ready guidance: install commands, a fully parameterized 'mythic-cli payload create' invocation, and a Python scripting example. Not a 5 because the body claims a 'JSON-RPC tasking model' while also listing a 'Direct GraphQL API' endpoint (an internal inconsistency), and the Python example's imports (mythic_utilities, mythic_callbacks) may not match the real mythic package's API. | 4 / 5 |
Workflow Clarity | The 9-step 'Common workflow' has a clear sequence but validation checkpoints are absent or only implicit (survey after callback, no verify-before-lateral-movement, persistence, or teardown steps). Since the workflow includes destructive and batch-style operations (SMB lateral to peers, persistence, cleanup/teardown), the rubric's cap of 3 for missing validation applies. | 3 / 5 |
Progressive Disclosure | Well-organized single-file skill (no bundle files exist) with clear section headers and navigable tables. Not a 5 because at ~130 lines the agent matrix, comparison table, and scripting API details are inline content that would fit better in one-level-deep reference files if the skill grows, and the 'References' section lists external docs rather than structured bundle navigation. | 4 / 5 |
Total | 16 / 20 Passed |