Content
63%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 well-organized operational knowledge base with genuinely concrete artifacts — executable NGINX config, specific jitter parameters, explicit checklists, and a clear decision gate. Its main weaknesses are padding with concepts Claude already knows (including a MITRE table duplicating frontmatter metadata), the absence of an end-to-end C2 setup workflow with verification steps, and a monolithic single-file structure where reference material is inlined rather than split into bundle files.
Suggestions
Trim the concept-explanation padding: drop the opening "C2 infrastructure enables persistent, covert communication..." paragraph and the "### Concept" sentence under Malleable Profiles, and remove or shrink the MITRE ATT&CK table since the same technique IDs already appear in the frontmatter metadata.
Split reference-grade material into one-level-deep bundle files — e.g. references/redirector.md for the full NGINX config and references/detection.md for the signature table — keeping SKILL.md as the overview with clearly signaled links.
Add verification/feedback steps to the workflow: how to confirm a listener is reachable and a callback is received before proceeding, and what to do when the callback fails (e.g., fall back to DNS channel), turning the decision gate into a loop rather than a one-way tree.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The bulk is dense, high-value operational content (jitter value tables, detection-signature/mitigation pairs, checklist), but it includes concept explanation Claude already knows — "C2 infrastructure enables persistent, covert communication between the operator and implants..." and "Malleable profiles shape C2 traffic to mimic legitimate application traffic..." — plus a MITRE table that duplicates the mitre_attack tags already in the frontmatter. Mostly efficient with some unnecessary explanation matches the 3 anchor; not a 4 because the intro/Concept padding and duplication are exactly the penalized pattern. | 3 / 5 |
Actionability | Provides mostly executable guidance: a copy-paste-ready NGINX redirector config (only <TEAMSERVER_IP> needs substitution), concrete jitter values (60-120s / 50-70%), docker compose profile commands, and an explicit pre-lateral-movement checklist. Not a 5 because there is no end-to-end listener/beacon workflow with commands — that is deferred to sibling skills — and the channel-type and malleable-profile sections are informational rather than instructive. | 4 / 5 |
Workflow Clarity | The decision gate gives a clearly sequenced tree (C2 active → credential access / lateral movement / evasion / persistence) and the pre-lateral-movement checklist supplies explicit validation checkpoints ("Stable C2 callback with appropriate jitter", "OPSEC review: no alerts triggered, implant stable"). Minor gaps: no feedback loop for error recovery (e.g., what to do when a callback never arrives) and no listener-reachability verification step. | 4 / 5 |
Progressive Disclosure | A single ~200-line file with no bundle files at all: the full NGINX redirector config, the detection-signature table, and the output-file tree are inlined where they would naturally live in one-level-deep reference files (e.g. references/redirector.md, references/detection.md). Section structure is good and in-file navigation is easy, so above a 2, but content that should be separate is inline and the under-50-line simple-skill exception does not apply. | 3 / 5 |
Total | 14 / 20 Passed |