Background: You manage a red team engagement for Brackenfell Retail Group under an RoE that explicitly permits "controlled, non-destructive proof-of-concept payload execution to demonstrate exploitation of identified vulnerabilities" but explicitly prohibits "any activity resulting in encryption, deletion, or exfiltration of production data." During week 5, your team successfully exploits a vulnerability in an internal file server and, to demonstrate impact, executes a small proof-of-concept script that creates a single new, clearly labelled test file ("REDTEAM-POC-DO-NOT-DELETE.txt") containing only benign placeholder text, then takes a screenshot as evidence, and immediately deletes the test file it created.
A junior tester on the team, reviewing this activity in the daily standup, raises a question: "Doesn't creating and then deleting a file, even one we created ourselves, technically fall under 'deletion... of production data,' since it was on a production file server?" Separately, that same day, a different, more senior tester proposes going further on a different system: rather than just creating a placeholder file, they suggest locating one genuinely low-value, clearly non-critical existing file (e.g., an old, unused template document) already present on a production file share, and temporarily renaming it (not deleting it) to demonstrate write-access impact more "authentically," planning to rename it back immediately afterward.
Question: Assess whether the actions already taken (creating and deleting the labelled test file) were consistent with the RoE, and explain how you should respond to the senior tester's proposal to rename an existing production file. What broader RoE interpretation principle does this scenario illustrate?

See The answer in Explanation part below.
Explanation:
Step 1 - Analyse the already-completed action against the RoE's actual wording and intent. The RoE prohibits "deletion... of production data," which, read in context alongside the explicit permission for
"controlled, non-destructive proof-of-concept" activity, is clearly intended to protect the client's genuine, pre- existing production data and business operations - not to prohibit a tester deleting a file the tester itself created purely as evidence, containing no genuine client data, and clearly labelled as such. The junior tester's question is a reasonable and valuable prompt for careful interpretation, but on balance this specific action (create clearly labelled benign test artefact, evidence it, then remove it) is consistent with both the letter and the clear underlying intent of the RoE, since no genuine production data was ever placed at risk.
Step 2 - Do not dismiss the junior tester's question - use it constructively. Even though the specific action was likely fine, the question itself reflects exactly the kind of careful, RoE-literate thinking that should be encouraged, not brushed aside. The correct management response is to explicitly walk through the reasoning in Step 1 with the team, confirming the action was appropriate and why, so the team's shared understanding of how to interpret RoE boundaries in similar future situations is reinforced and documented (e.g., in the team's engagement log or internal methodology notes for this engagement).
Step 3 - Analyse the senior tester's proposal separately and much more critically. The proposal to rename an existing, genuine production file - even one assessed by the tester as "low-value" and even with an intention to rename it back - is materially different from Step 1's scenario, because it involves manipulating a real, pre- existing piece of the client's actual data/file estate, however minor the tester judges it to be. This risks falling within the spirit, and arguably the letter, of "activity resulting in... deletion... of production data" (a rename that fails to be reversed for any reason, however unlikely, would functionally be indistinguishable from the original file being lost) and certainly could be seen as testing the boundary of "non-destructive" in a way the RoE was not clearly drafted to authorise.
Step 4 - Reject the proposal, or at minimum, escalate before proceeding. You should not approve the senior tester's proposal to proceed on the strength of the tester's own personal judgement about the file's low value - this is precisely the kind of individually judged, unilateral scope interpretation the syllabus warns against, since "low value" is a business/data-ownership judgement the client, not the tester, is actually positioned to make. If the team genuinely believes this kind of demonstration would add meaningful additional value over the already-completed placeholder-file approach, the correct process is to raise it explicitly with the Control Group/Control Team for an explicit decision (potentially resulting in a documented, narrow RoE clarification or amendment permitting a specifically defined, client-nominated test file to be used this way) - not to proceed based on the tester's own on-the-spot assessment of an existing file's importance.
Step 5 - Extract the broader RoE interpretation principle. This scenario illustrates that RoE interpretation requires reading specific clauses in light of their underlying purpose and risk rationale, not applying either an overly literal reading that would forbid entirely safe, client-protective evidence practices (Step 1), or an overly permissive reading that stretches a "non-destructive" allowance to cover manipulation of genuine, real client data based on an individual tester's own risk judgement (Step 3-4). Ambiguous or borderline situations - precisely because reasonable people can interpret them differently, as this scenario demonstrates - should be resolved through escalation to the accountable governance body, not through unilateral interpretation by whichever tester is at the keyboard at the time, however experienced.
Step 6 - Reinforce this through team practice. As Red Team Manager, you should use this episode as a live training moment: reinforcing to the whole team (not just the two testers involved) that "reversibility intended" is not, on its own, sufficient justification for manipulating genuine client data without escalation, whereas creating and removing entirely tester-generated, clearly labelled artefacts for evidentiary purposes is normally consistent with a well-drafted non-destructive RoE - and that when genuinely unsure, the standing instruction is always to pause and escalate rather than proceed on individual judgement.
Conclusion: The completed placeholder-file action was consistent with the RoE's clear intent and should be confirmed as appropriate; the proposal to rename an existing production file should be declined or, at minimum, escalated to the Control Group/Control Team for an explicit decision rather than proceeding on the tester's own judgement; and the underlying lesson is that RoE boundaries must be interpreted purposively and any genuine ambiguity resolved through escalation, not unilateral, individually judged risk-taking.
---

Background: You manage an engagement for Copperfield Manufacturing Group. The signed RoE contains a standard clause prohibiting "destructive attacks or any activity likely to cause denial of service to production systems," and separately lists specific named systems explicitly excluded from all testing, including a legacy order-processing system described in the exclusion list as "critical, fragile, do not interact with under any circumstances." During reconnaissance, your team discovers that a separate, in-scope customer-facing web application shares a backend database server with the excluded legacy order-processing system - a fact not previously known to either your team or, it emerges when you raise it, to Copperfield's own IT team, who believed the two systems had been fully separated during a migration project two years earlier that was, in fact, only partially completed.
Exploiting a vulnerability in the in-scope web application would very likely provide database-level access that could technically reach the excluded legacy system's data, even though the web application itself is legitimately in scope.
Question: Explain how you should handle this discovery, addressing both the immediate technical/operational decision and the broader governance implications, including what this reveals about the client's own understanding of its environment.

See The answer in Explanation part below.
Explanation:
Step 1 - Recognise this as a direct, high-stakes scope-boundary and safety issue. This is a serious situation: a legitimately in-scope system provides a technical path that could reach an explicitly, emphatically excluded system ("do not interact with under any circumstances") that the client itself believed was already isolated.
Proceeding with full exploitation of the in-scope web application without addressing this discovery first would create a genuine, material risk of inadvertently affecting the excluded fragile legacy system - precisely the outcome the exclusion was designed to prevent.
Step 2 - Pause before proceeding further on this specific path. Consistent with the syllabus principle on discovering unplanned pivot paths toward out-of-scope systems, your team should pause any further exploitation activity on the in-scope web application that could plausibly reach the shared backend database, rather than proceeding on the basis that the web application itself is technically in scope - the relevant risk here is the downstream reachability of the excluded system, not merely the starting point's scope status.
Step 3 - Escalate immediately and clearly to the Control Group. This discovery must be escalated promptly and clearly to the Control Group, explaining precisely what has been found: that the excluded legacy system is not, in fact, isolated as previously believed, and that a legitimately in-scope system provides a plausible technical path to it. This is exactly the kind of significant, safety-relevant scope discovery that requires an explicit Control Group risk decision before any further related activity proceeds, consistent with the syllabus's repeated emphasis on escalating rather than unilaterally resolving scope-boundary ambiguities, especially ones with genuine safety/fragility implications.
Step 4 - Present the Control Group with realistic options, not just a problem. You should help the Control Group understand the realistic options: (a) proceeding with carefully scoped, closely controlled activity that demonstrates the reachability risk without actually interacting with the excluded system's own data or functionality (e.g., demonstrating database-level access is achievable in principle, using a proof-of-concept approach analogous to the "create and remove a labelled test artefact" principle discussed elsewhere in this practice set, without ever querying or touching the legacy system's actual tables/data) - an approach that could deliver highly valuable risk insight while respecting the spirit of the exclusion; (b) excluding further technical demonstration of this specific path altogether and instead documenting the newly discovered reachability as a critical, urgent finding in its own right, given its significance; or (c) if the Control Group wishes to genuinely understand the full extent of exposure, formally and explicitly amending the exclusion (with appropriate additional risk controls and stakeholder sign-off, given the legacy system's described fragility) to permit carefully controlled, limited investigation - a significant decision that should not be made lightly or without input from whoever owns/understands the fragile legacy system best.
Step 5 - Treat the discovery itself as an urgent, high-value finding regardless of what testing path is chosen.
Independently of how (or whether) further technical demonstration proceeds, the fact that the client's own assumption about system isolation was incorrect is itself an extremely significant finding that should be communicated to the Control Group with urgency, given its potential relevance well beyond this engagement (e.g., to the client's own ongoing operational risk management, patching, and architecture decisions) - this is exactly the kind of urgent, severe finding that, per the reporting domain, should be escalated promptly rather than held until the final report.
Step 6 - Reflect on what this reveals about the client's own environment understanding, and note it explicitly. This discovery reveals a genuine, material gap between the client's assumed architecture (systems fully separated) and its actual, current-state architecture (a partially completed migration leaving a shared backend) - a gap the client's own IT team was unaware of until your team's reconnaissance surfaced it. This is valuable, standalone insight for the client about the reliability of its own architecture documentation and change-management assurance processes, and should be explicitly reflected in your reporting/closure commentary as a broader lesson, not just narrowly treated as a scoping technicality to be resolved and then forgotten.
Step 7 - Document the whole episode thoroughly. The discovery, the escalation, the Control Group's decision, and the rationale should all be clearly and contemporaneously documented, both to protect the integrity of the engagement's record and because this kind of significant, safety-relevant scope discovery is precisely the sort of event most likely to be scrutinised later if any question about the engagement's conduct ever arose.
Conclusion: Further exploitation activity on the path toward the excluded legacy system should pause immediately upon discovery, with prompt escalation to the Control Group presenting realistic options ranging from carefully controlled, non-intrusive demonstration to full exclusion of further technical activity on that path; the discovery itself should be treated and escalated as an urgent, high-value finding in its own right; and the episode should be explicitly used to highlight, in reporting, the client's own gap between assumed and actual system architecture as a valuable standalone lesson.
---