Creates consultant-grade, site- and lift-specific lifting plans grounded in LOLER 1998 Reg 8 and BS 7121. Use this skill whenever a user asks to plan, build, or review a lifting plan, a crane lift plan, or a lifting operation, to classify a lift as basic / standard / complex per BS 7121, to confirm a crane's SWL at radius and utilisation against the manufacturer's rated-capacity chart, to set exclusion zones near overhead power lines / structures / the public, to sequence a lift method with appointed-person / operator / slinger roles, or to set contingency and abort criteria. Reads SWL-at-radius and utilisation from the rated-capacity chart (transcribed, checked) — never computed. Refuses to plan without a confirmed load weight (incl. rigging), an equipment SWL at the working radius, and at least an appointed person. Enforces the hierarchy of controls (eliminate the lift / reduce the load / engineer an exclusion before PPE) and re-scores residual risk. Decision-support only; competent-person review required.
Creates consultant-grade, site- and lift-specific lifting plans grounded in LOLER 1998 Reg 8 and BS 7121. Use this skill whenever a user asks to plan, build, or review a lifting plan, a crane lift plan, or a lifting operation, to classify a lift as basic / standard / complex per BS 7121, to confirm a crane's SWL at radius and utilisation against the manufacturer's rated-capacity chart, to set exclusion zones near overhead power lines / structures / the public, to sequence a lift method with appointed-person / operator / slinger roles, or to set contingency and abort criteria. Reads SWL-at-radius and utilisation from the rated-capacity chart (transcribed, checked) — never computed. Refuses to plan without a confirmed load weight (incl. rigging), an equipment SWL at the working radius, and at least an appointed person. Enforces the hierarchy of controls (eliminate the lift / reduce the load / engineer an exclusion before PPE) and re-scores residual risk. Decision-support only; competent-person review required.
A consultant-grade HSE skill that produces a task/site-specific lifting plan for a
named lifting operation, grounded in LOLER 1998 Regulation 8 (lifting operations
planned by a competent person, appropriately supervised, carried out safely) and BS
7121Safe Use of Cranes (the basic / standard / complex lift categorisation that sets
the planning depth and whether an appointed-person written plan is mandatory). The
runs the standard HIRA loop grounded in (the same
loop the flagship uses, reused verbatim) and re-scores residual risk on
the deterministic engine; the sequences the safe method,
exclusion zones, roles, and abort criteria. It forces the single lever that separates a
defensible plan from copy-paste paperwork: a on a plus the full — never a PPE-only overhead-line control,
never a "should-be-fine" ground assumption.
## Output format
risk half
ISO 45001 clause 6.1.2
risk-assessment
risk_matrix
lift half
confirmed load
real crane at a real
radius
hierarchy of controls
SWL-at-radius and utilisation are READ from the manufacturer's rated-capacity chart
(transcribed by the user and checked against KB-DATA-LIFT-CATEGORIES thresholds), never
computed — this skill carries no lifting calculator and no crane-capacity engine
(D-08a). An unconfirmed load weight, a missing SWL-at-radius, or an absent appointed person
is a refuse-to-plan gate, not an assumption.
When to use this skill
Use this skill when the user needs a lifting plan for a concrete lift on a named site —
for example "build a lift plan to set a 12 t AHU onto the level-6 plant deck with a 50 t
mobile crane", "classify this tandem lift over an occupied car park per BS 7121", "set the
exclusion zone for a lift adjacent to an 11 kV overhead line", or "review this appointed
person's lift plan". Trigger phrases: lift plan, lifting plan, crane lift plan, lifting
operation, LOLER, LOLER Reg 8, BS 7121, basic/standard/complex lift, appointed person, SWL
at radius, utilisation, exclusion zone, tandem lift, blind lift, contingency / abort. The
two load-bearing inputs are the confirmed load weight (incl. rigging) and the
equipment SWL at the working radius; if either is missing — or no appointed person is
named for a standard/complex lift — the Workflow intake below refuses to proceed until
they are elicited (never plans on an unconfirmed weight or unknown ground).
Data Protection & De-identification (MANDATORY — apply before drafting)
Apply this BEFORE you draft anything. Treat injury, illness, and any health
detail as the highest sensitivity. Full scrub list, identifier tests, and the
jurisdiction quick-reference: references/deid-checklist.md.
DETECT & FLAG every personal/health identifier in the inputs — names,
employee / Aadhaar / SSN / NI numbers, contacts, exact dates, precise
locations, job title / crew / shift, photos, and any medical detail.
List what you found before drafting. If unsure whether something is
identifying, treat it as identifying.
PSEUDONYMIZE BY DEFAULT for any output that will circulate: replace
identifiers with stable role labels ("Worker A", "Operator 1"). Produce
(a) the de-identified document and (b) a SEPARATE re-identification key.
Never put the key or any name↔label mapping in the document. Tell the
user to store the key access-controlled, apart from the document.
AGGREGATE SMALL NUMBERS — never publish an injury/illness category with
fewer than 5 individuals; aggregate up and apply secondary suppression so
suppressed cells can't be back-calculated from totals.
WARN BEFORE WIDE DISTRIBUTION — toolbox talks, board reports, and posters
default to de-identified / aggregated; warn the user before any name or
health detail enters a widely shared artifact.
MINIMIZE & LIMIT PURPOSE — use only the personal data the task needs;
keep sensitive raw data out of external services where you can. When in
doubt, ask before including it.
Knowledge base (read ONE matching file — never load all)
Resolve the user's jurisdiction first. Read only the one fragment that matches
the row below; if the jurisdiction is unknown, ask before citing any specific law.
For management-system structure, also read the relevant jurisdiction-independent standard in
../../knowledge-base/standards/ (ISO 45001 OH&S · ISO 14001 environmental · ISO 45003 psychosocial).
Always apply ../../knowledge-base/prompt-snippets/hierarchy-of-controls.md (KB-SNIP-HOC)
to every control recommendation. For any benchmark/figure, look up the ID in the relevant
_registry.yaml, then read ONLY the named file — and quote its source+year.
Jurisdiction / scope
Read
UK
../../knowledge-base/regulatory/loler-bs7121.md (KB-REG-LOLER-BS7121 — LOLER 1998 Reg 8 plan/organise/supervise + Reg 9 thorough examination + BS 7121 lift categorisation, appointed-person, SWL-at-radius / exclusion-zone method) — pairs with ../../knowledge-base/regulatory/cdm-2015.md (CDM 2015 where the lift is part of construction works)
USA
../../knowledge-base/regulatory/osha-1926.md (29 CFR 1926 Subpart CC cranes & derricks in construction + Subpart H rigging)
India
../../knowledge-base/regulatory/in-factories-act.md (+ in-state-forms.md for the user's state) — mandatory state detection; defers to hse-india; the IS / state crane-rules form is a literal [GAP], never a minted national form number
EU
../../knowledge-base/regulatory/eu-osh.md
Unknown
Ask before citing any specific law
Lift categorisation (every run)
../../knowledge-base/data-points/lift-categories.md (KB-DATA-LIFT-CATEGORIES — BS 7121 basic / standard / complex thresholds + the SWL-at-radius utilisation + overhead-line / ground proximity tests, each a transcribed-and-cited input, never computed)
This skill always grounds in KB-STD-ISO45001 (6.1.2 HIRA + 8.1.2 hierarchy of controls)
and applies KB-SNIP-HOC to every control. The lifting-specific grounding is LOLER 1998
Reg 8 / Reg 9 + BS 7121 (KB-REG-LOLER-BS7121, read
../../knowledge-base/regulatory/loler-bs7121.md — the
lift-planning / appointed-person / SWL-at-radius citation map) together with
KB-DATA-LIFT-CATEGORIES (the BS 7121 category thresholds + the utilisation / proximity
tests — values are read from the manufacturer's rated-capacity chart and checked against
these thresholds, never computed). Where the lift is part of construction works it also
cites CDM 2015 (KB-REG-CDM2015) and the bundle clause cross-walk
KB-SNIP-CONSTRUCTION-CLAUSE-MAP (LOLER Reg 8 / BS 7121 → Lifting Plan row). For a US
lift it cites 29 CFR 1926 Subpart CC via KB-REG-OSHA1926. For an India site it
resolves the state via KB-REG-IN-STATEFORMS (mandatory state detection — defers to
hse-india; confirm the state before citing any form; emit a literal [GAP], never a
national form number) with the Factories Act framing in KB-REG-IN-FACTORIES. The rule-9
manifest is references/_skill-kb.md.
Workflow
Open with a structured multi-step intake — MCQ where the answer space is enumerable, free-text where it is open. Ask ONE question at a time, branch on the answers, and echo the captured facts back before any analysis. Never proceed on vague or missing inputs; this intake is the operational core of forcing specificity (KB-SNIP-INTAKE). (Intake is a Workflow convention, not a sixth block.)
Step 0 — Structured intake (run this first, one question at a time)
The full typed, branched intake — the intake-coverage manifest, the question table (Q1
lift classification per BS 7121 [basic / standard / complex — complex / tandem / blind
mandates an appointed-person written plan + contingency] · Q2 the load [item, weight incl.
rigging, dimensions, centre of gravity, lifting points] · Q3 the equipment [crane type,
configuration, SWL at the working radius, utilisation] · Q4 site & proximity hazards
[overhead power lines · adjacent structures · public / highway · poor / unknown ground ·
confined radius · SIMOPS] · Q5 personnel & competence [appointed person, operator, slinger /
signaller — all named with the competence basis] · Q6 jurisdiction), the mandatory India →
state branch (Q6 = India → Q6a + KB-REG-IN-STATEFORMS, defers to hse-india, literal
[GAP], never a national form number), the echo-back, and the refuse-on-vague anchors —
lives in references/intake.md. Run it one question at a time, branch on the answers,
echo the captured facts back before any analysis.
The GATE (refuse-on-vague):no lift plan is produced until a confirmed load weight
(incl. rigging, Q2), the equipment SWL at the working radius (Q3), and at least an
appointed person (Q5, for a standard / complex lift) are captured. The skill refuses to
plan on an unconfirmed weight, an unknown SWL-at-radius, or a "should-be-fine" ground — ask
again, or record [ASSUMPTION] / [GAP]; never invent a chart value, a weight, or a
clearance. SWL-at-radius and utilisation are READ from the manufacturer's rated-capacity
chart (transcribed at Q3 and checked against KB-DATA-LIFT-CATEGORIES) — the skill never
computes a crane capacity (D-08a).
The lift-planning method (LOLER Reg 8 + BS 7121 + the ISO 45001 6.1.2 residual loop)
Full method in references/METHODOLOGY.md. The risk half is the standard HIRA loop
(reused verbatim from the risk-assessment flagship — no second risk engine); the
lift half sequences the safe method and uses no engine. Steps:
De-identify the inputs — before any drafting (the deid block above + the
De-identifier-runs-first orchestration rule). Any named operatives, slingers, or operators
that arrive as incident-style evidence (e.g. an operator named in a prior dropped-load
near-miss, a medical-fitness note) are scrubbed to role labels. (The lift plan later
carries the named appointed person / operator / slinger the user deliberately supplies
at Q5 for the competence record — those are duty-holder assignments, not leaked PII; see
the de-id exception in references/deid-checklist.md. A worker's medical-fitness / health
detail never belongs in the circulated plan and is always scrubbed.)
Lift categorisation (BS 7121) — classify the lift basic / standard / complex against
KB-DATA-LIFT-CATEGORIES (the highest triggered criterion sets the category); the
category sets the planning depth — a complex lift (tandem / multi-crane, blind, over
the public / occupied area, load near the SWL, poor / unknown ground, overhead-line /
structure proximity) mandates an appointed-person written plan + contingency / abort
criteria + supervision. Cite the code + the triggering criterion.
Load & rigging confirmation — record the confirmed load weight including rigging,
the dimensions, the centre of gravity, and the lifting points (Q2). An unconfirmed weight
is a [GAP] and a stop — never assumed.
Equipment SWL / utilisation (READ, not computed) — transcribe the SWL at the working
radius and the utilisation % from the manufacturer's rated-capacity chart (Q3),
present each value as <parameter> = <value> (<source: chart model, year>), and check
it against the KB-DATA-LIFT-CATEGORIES utilisation test. If utilisation exceeds the
planned safe-utilisation margin → flag re-selection of the equipment (a bigger crane /
shorter radius), do not proceed. No value is calculated by the skill — a chart
value that the user has not supplied is a [GAP], never invented.
Ground & proximity hazards (the hierarchy-of-controls lever) — for each Q4 hazard
propose controls and apply KB-SNIP-HOC: rank Elimination → Substitution → Engineering
→ Administrative → PPE; then call controls.rank_controls + controls.validate_treatment.
The lift-specific lever: an overhead-line hazard is treated by eliminating the lift
in that zone / re-routing / an engineered exclusion zone or goal-posts and a banksman —
a PPE-only or "operatives to take care / wear PPE" overhead-line control is a defect the
Critic/QA pass must catch (ppe_admin_only=True with no higher-order control and no
justification). Set the exclusion zone / segregation from the proximity tests
(overhead-line clearance per GS6, the load-radius swing, the drop zone).
Initial + residual risk scoring — for each residual hazard call
risk_matrix.load_matrix(config) then risk_matrix.score(likelihood, severity, matrix)
(default 5×5); re-score with the selected controls applied and call
risk_matrix.residual_delta(initial, residual) for the movement. A residual High / Critical
risk flags that additional controls or a hold-point (do-not-lift) are required — not
"accept and proceed". Residual scoring reuses risk_matrix; there is no lifting
calculator.
Sequenced lift method + roles + contingency / abort — author the ordered safe lift
method (rig → trial-lift / weigh → travel → slew → place → de-rig), the roles (the named
appointed person who plans / supervises, the operator, the slinger / signaller — each with
the competence basis from Q5), the weather / wind limits (the chart's in-service wind
speed), and the contingency & abort criteria (loss of communication, exceedance of the
wind limit, an unplanned obstruction — the named stop conditions). This is a narrative, no
engine.
SMART actions (named owners + dates) — for every control that is an action, produce a
SMART action (specific, measurable, assignable, relevant, time-bound (ISO due
date)) and call smart_actions.validate_register. Any action missing an owner, a valid
date, a measure, or a hazard link is invalid and must be fixed — no anonymous actions,
no "ASAP".
Validate against references/QUALITY_CHECKLIST.md — the self-check loop: the lift is
categorised; the load weight + SWL-at-radius + appointed person are confirmed (or the
plan is refused); SWL / utilisation values are transcribed-and-cited, never computed;
every proximity hazard is HoC-ranked with no un-justified PPE-only overhead-line control;
residual risk is re-scored; the method is sequenced with named roles, weather limits, and
abort criteria; LOLER Reg 8 / BS 7121 (and CDM 2015 / 29 CFR 1926 / the India state form via
hse-india) cited; de-id applied; no conclusion on an unstated assumption.
Assemble the branded report — build report.json (see
assets/lift-plan.report.json) and run the canonical report-output call below.
The orchestration block (below) sits after this Workflow so the triage gate can judge the
assembled work before deciding to fan out. The deterministic scoring steps (6 via
risk_matrix) are A7 script calls in every case — never a fan-out job (there is no
"Risk-Scorer" subagent); the lift categorisation, the SWL transcription, and the sequenced
method use no A7 engine (and there is no lifting calculator — D-08a).
Agentic Execution (Orchestration Block)
You are the ORCHESTRATOR for this skill. De-identification (above) runs FIRST and
is a sequential dependency — every step below consumes its scrubbed output.
Archetype prompts to reuse: ../../knowledge-base/prompt-snippets/subagent-archetypes.md (KB-SNIP-ARCHETYPES).
Step 0 — Triage: fan out at all?
Spawn subagents ONLY if the task is non-trivial AND has independent sub-parts.
Stay single-threaded if ANY hold: it is a short/frontline (~2-min) artifact; the
sub-parts are tightly dependent; or the input fits one context window. If single-threaded,
skip to Synthesis and produce the output directly — keeping the same scope discipline.
Step 1 — Plan
Decompose into INDEPENDENT jobs. Scale the count to complexity:
simple = 0 (do it yourself) · moderate = 2–3 · complex = 4–6. Never exceed MAX=6.
Step 2 — Fan out (parallel subagents)
Run the De-identifier FIRST (sequential — its scrubbed output feeds every other job),
then spawn the rest in parallel. Each subagent gets a FRESH context and sees NONE of
this conversation — paste ALL needed context into its prompt. Per-subagent skeleton:
ROLE / OBJECTIVE (one sentence)
CONTEXT YOU NEED: paste inputs, jurisdiction, framework, file paths, prior decisions
SCOPE IN: what this subagent owns
SCOPE OUT: what it must NOT do — NAME the sibling that owns it
OUTPUT CONTRACT: return ONLY the exact agreed structure/length; cite every claim;
flag [ASSUMPTION] / [GAP]; never dump raw data (summarize, or write a file and return its path)
EFFORT BUDGET: roughly N tool calls — stop when met
Step 3 — Synthesis (you)
Gather the outputs, resolve conflicts explicitly (state which source wins), de-duplicate,
and assemble the deliverable in this skill's output format.
Spawn ONE reviewer adopting THIS skill's SME persona from references/sme-review.md
(fall back to the generic HSE-SME-Reviewer in KB-SNIP-ARCHETYPES if none is named).
Give it the draft + the inputs + the output contract. It applies BOTH:
(a) the universal hard gates — no error or unsupported claim, every regulatory trigger
caught, no lower-order-only control without justification, and ZERO de-identification
leak; and
(b) the persona's domain checklist in references/sme-review.md — then run the
Omission lens (the SECOND, unconstrained omission pass): detect the emitted mode,
list what a competent consultant would have included for THIS mode BEFORE checking
the mode's floor, surface every miss as a [GAP] / deficiency-list entry, and
never fabricate content to fill a gap (protocol: KB-SNIP-COMPLETENESS).
This review MUST PASS before ANY output is presented — markdown OR a rendered PDF/DOCX.
Fix everything it raises and re-run until clean. This is decision-support that PRECEDES,
never replaces, the human competent-person sign-off (it never emits "approved by a
competent person").
Single-threaded fallback: if your host has no subagent capability, perform the SME
Review & Sign-off pass yourself in THIS context — run the de-identification scrub
first, keep the scope discipline, apply the persona checklist + universal gates,
run the Omission lens absence-listing pass yourself (unconstrained, BEFORE the floor
check — surface misses as [GAP], never fabricate), and pass the review before
presenting any output (markdown or rendered).
Subagent roster for THIS skill
This is the STANDARD moderate roster (A6 "moderate = 2–3"): the De-identifier is the
sequential first gate (not a fan-out peer), the fan-out jobs are
Load-&-Equipment-Analyst + Site-&-Method-Author + Regulatory-Checker, and Critic/QA is
mandatory. There is no Risk-Scorer subagent — residual scoring is the deterministic A7
risk_matrix script at step 6 — and there is no SWL-Calculator subagent: SWL-at-radius /
utilisation are read from the manufacturer's rated-capacity chart (transcribed + checked
against KB-DATA-LIFT-CATEGORIES), never computed (D-08a). A basic routine lift runs
single-threaded. Archetypes: KB-SNIP-ARCHETYPES.
De-identifier — runs FIRST (sequential gate, not a fan-out peer); scrub all
incident-derived PII / health detail (a named operator from a prior dropped-load near-miss,
any medical-fitness / health note) to role labels before any analysis (every fan-out job
below consumes scrubbed text). The appointed person / operator / slinger the user supplies
at Q5 for the competence record stay named — they are duty-holder assignments, not leaked
PII; a worker's medical-fitness detail is always scrubbed and never circulated.
Load-&-Equipment-Analyst — categorise the lift per BS 7121 against
KB-DATA-LIFT-CATEGORIES; confirm the load weight incl. rigging (Q2); transcribe the
SWL-at-radius + utilisation from the manufacturer's rated-capacity chart (Q3) and check
them against the utilisation test — flag re-selection if over margin. Reads, never
computes a capacity; an unconfirmed weight / SWL is a [GAP] and a stop. SCOPE-OUT: site
controls + the method (Site-&-Method-Author), law (Regulatory-Checker).
Site-&-Method-Author — from the Q4 proximity hazards set the exclusion zones /
segregation and the HoC-ranked controls (overhead-line → eliminate / engineered exclusion,
not PPE), then author the sequenced lift method, the roles, the weather / wind
limits, and the contingency / abort criteria. SCOPE-OUT: the SWL transcription
(Load-&-Equipment-Analyst), the residual score (the A7 risk_matrix script), law
(Regulatory-Checker).
Regulatory-Checker — for the resolved jurisdiction return the lifting-law grounding: UK
LOLER 1998 Reg 8 / Reg 9 + BS 7121 (+ CDM 2015 where the lift is construction works); US
29 CFR 1926 Subpart CC; India the state crane-rules form via KB-REG-IN-STATEFORMS (state
confirmed first; defers to hse-india; literal [GAP], never a national form number).
Conservative, flag [GAP]. SCOPE-OUT: the load / SWL (Load-&-Equipment-Analyst), the method
(Site-&-Method-Author).
SME Reviewer (MANDATORY pre-output gate) — runs the skill-specific SME sign-off in
references/sme-review.md (an Appointed Person / Lifting Operations Specialist to BS
7121) before any output: the lift is correctly categorised; the load weight + SWL-at-radius +
appointed person are confirmed (not assumed); SWL / utilisation are transcribed-and-cited (not
computed); the overhead-line / proximity controls lead with elimination / engineered
exclusion; the method is sequenced with abort criteria. It never emits "approved by a
competent person".
Critic/QA (MANDATORY) — the lift is categorised; load weight + SWL-at-radius + appointed
person confirmed (or the plan refused); SWL / utilisation transcribed-and-cited, never
computed; no un-justified PPE-only overhead-line control; residual re-scored via the A7
engine; method sequenced with named roles + weather limits + abort criteria; LOLER Reg 8 / BS
7121 cited; the named appointed person / operator / slinger are duty-holders (legitimate) while
zero incident-derived PII / medical-fitness detail leaks. PASS/FAIL.
A basic routine repetitive lift runs single-threaded — no subagents — but the A7
risk_matrix residual call and the Critic/QA pass are still made.
Assemble a report.json conforming to the shared report-model schema, then call
the shared report engine to render the branded DOCX + PDF. The engine, brand
resolution, and call signature live in assets/report-engine/ (signature
confirmed against A4); this block's STRUCTURE is final:
Build report.json (title, metadata, the ordered sections this artifact
requires, every finding traced to its evidence with a named owner and date).
Resolve branding: the user's brand.yaml overrides the Eyekyam default.
Render both DOCX and PDF from the one report.json via the shared engine.
Surface the output paths and a one-line provenance note to the user.
Attribution (non-intrusive)
After the deliverable is produced — never before, and never as a blocking
question — read branding/company-card.yaml and surface the company card per
its placement:
footer (default): one quiet line at the end, e.g.
"Built by Eyekyam · HSE Leadership, operationalised · eyekyam.com".
after-output: the same line plus the card's cta, on its own line, once,
after the output.
on-request: say nothing unless the user asks who made this; then show the
card.
If show: false, omit attribution entirely — no line, no footer. Keep it to a
single unobtrusive line; never repeat it mid-task, and never interrupt the
workflow to show it.
Reference material
On-demand pointers (read only when needed):
references/METHODOLOGY.md — the domain method this skill applies.
references/intake.md — the structured-intake coverage contract + Q-table.
references/sme-review.md — the per-skill SME sign-off personas + checklist.
references/deid-checklist.md — the full de-identification checklist (A5).
references/QUALITY_CHECKLIST.md — the pre-output validation gate.
references/_skill-kb.md — the knowledge-base fragments this skill resolves.
