Browse all skills & agents

manual-testing-overview

Overview of manual testing for testers new to it: explains the two modes (session-based exploratory and scripted manual), lays out a recommended starting path, and names the four heuristic references (SFDPOT, HICCUPPS-F, FCC CUTS VIDS, CRUSSPIC STMPL) with guidance on when to reach for each. Use as a single entry point when getting started with manual or exploratory testing.

Install with skills.sh (any agent)

npx skills add testland/qa --skill manual-testing-overview

manual-testing-overview

Overview

There are two modes of human-driven testing. Session-based exploratory testing (SBTM - Jonathan and James Bach, 2000) is a framework in which a tester executes a focused, time-boxed session (60-90 minutes) guided by a charter - a mission statement of the form "Explore <area> with <tools> to discover <information>" - and records findings in a session sheet that a lead can review and debrief; the framework turns ad-hoc clicking into a measurable, chartered, and reviewable activity documented at satisfice.com/sbtm. Scripted manual testing follows the opposite pattern: a pre-authored step-table or Gherkin script tells the tester exactly what to do, what data to use, and what outcome to expect; the tester's job is execution and observation, not design. Both modes belong in a mature testing strategy - exploratory sessions surface unexpected bugs and learn the product; scripted tests give stakeholders a repeatable, signed-off record of verified behaviour.

Starting path: exploratory testing

If you want to run an exploratory session, follow this sequence:

Write a session charter. State the feature and the risk areas you care about, then draft a ready-to-execute mission card: a charter statement, scoped areas, applicable HICCUPPS-F oracles, SFDPOT elements to vary, and 1-3 tours to try. The four heuristic references below supply the vocabulary - you do not need to memorize them in advance.
Run the session using exploratory-tours-reference heuristics. Open the skill during your session to pick a tour if you feel the exploration drifting. The seven canonical Whittaker tours (Feature, Money, Landmark, Intellectual, Bad-data, Configuration, Garbage collector's) each frame the session around a different class of bugs. Pick the tour your charter hints at or the one that matches the risk you are currently investigating.
Debrief the session. Review your session sheet against every PROOF field (Past, Results, Obstacles, Outlook, Feelings) for completeness: flag a thin Feelings entry, watch for a skewed TBS time split (setup% > 30% signals an environment problem), and pick the next charter from your Outlook section.

Starting path: scripted testing and UAT

For scripted manual testing or UAT - where a human tester or business stakeholder needs a step-by-step script rather than an open-ended mission - start with manual-test-script-author. Give it a feature spec and it emits either a step-table (preconditions, steps, expected result, actual, pass/fail, notes) or Gherkin (Given/When/Then) format, self-contained with all required data, one scenario per script. For stakeholder sign-off scenarios, see uat-script-author which formats the same content in business language with explicit acceptance-criteria verification and a sign-off block.

The four heuristic references

These four skills are pure-reference catalogs. Read the one that matches your situation - knowing when each applies helps you ask the right question during a session.

Heuristic	What it catalogs	Reach for it when...
HICCUPPS-F (`hiccupps-f-heuristic`)	Michael Bolton's eight oracle types for deciding "is this a bug?" (History, Image, Comparable products, Claims, Users' desires, Product consistency, Purpose, Standards, Familiar problems)	You observe something and are not sure whether it is a defect - walk each oracle type to find the expectation it violates
SFDPOT (`sfdpot-exploratory-heuristic`)	James Bach's six dimensions of what to vary in a system (Structure, Function, Data, Platform, Operations, Time)	Your charter is set but you want to make sure your session varies the system in more than one dimension - SFDPOT is the "what to poke" checklist
FCC CUTS VIDS (`fcc-cuts-vids-heuristic`)	James Bach's mnemonic for modelling the system under test - Functions, Content, Criteria, Configurability, Users, Testability, States, Variability, Interfaces, Data, Scenarios	You are authoring a charter for an unfamiliar feature and need a systematic way to decompose what the system IS before deciding what to explore
CRUSSPIC STMPL (`crusspic-stmpl-heuristic`)	James Bach's twelve quality-criteria heuristic for characterising what "good enough" means (Capability, Reliability, Usability, Security, Scalability, Performance, Installability, Compatibility, plus Supportability, Testability, Maintainability, Portability, Localizability)	You are writing a test strategy or charter and need to decide which quality dimensions apply - CRUSSPIC STMPL is the "what quality means here" checklist

What to do first

If you are completely new: pick any feature you are currently working on and write a charter for it (Step 1 above). The charter names the heuristics your session should apply. Read the relevant heuristic skill once to understand the vocabulary, then run the session. File your session sheet and debrief it (Step 3). After two or three cycles the vocabulary will be natural.

References

Bach J., Bach J. Session-Based Test Management (2000) - satisfice.com/sbtm.
Whittaker J. Exploratory Software Testing (Addison-Wesley, 2009) - source of the seven canonical tours in exploratory-tours-reference.
Bolton M. HICCUPPS-F oracle heuristic - developsense.com.
Sibling skills: session-based-test-management-reference, exploratory-tours-reference, hiccupps-f-heuristic, sfdpot-exploratory-heuristic, fcc-cuts-vids-heuristic, crusspic-stmpl-heuristic, manual-test-script-author, uat-script-author.